Disease Surveillance for Malaria

7/31/2019 Disease Surveillance for Malaria

http://slidepdf.com/reader/full/disease-surveillance-for-malaria 1/84

Diseasesurveillance

forMalaria

control

a p m





Diseasesurveillance

for

Malaria

control

a p m



WHO Library Cataloguing-in-Publication Data

Disease surveillance or malaria control.

1.Malaria - prevention and control. 2.Epidemiologic surveillance. 3.Communicable disease

control. 4.Disease vectors. 5.Case management. I.World Health Organization.

ISBN 978 92 4 150334 1 (NLM classication: WC 765)

© World Health Organization 2012

All rights reserved. Publications o the World Health Organization are available on the WHO web site (www.who.int) or can bepurchased rom WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; ax:+41 22 791 4857; e-mail: [email protected]).

Requests or permission to reproduce or translate WHO publications – whether or sale or or noncommercial distribution – shouldbe addressed to WHO Press through the WHO web site (http://www.who.int/about/licensing/copyright_orm/en/index.html).

Te designations employed and the presentation o the material in this publication do not imply the expression o any opinion

whatsoever on the part o the World Health Organization concerning the legal status o any country, territory, city or area or o itsauthorities, or concerning the delimitation o its rontiers or boundaries. Dotted lines on maps represent approximate border linesor which there may not yet be ull agreement.

Te mention o specic companies or o certain manuacturers’ products does not imply that they are endorsed or recommended by the World Health Organization in preerence to others o a similar nature that are not mentioned. Errors and omissions excepted,the names o proprietary products are distinguished by initial capital letters.

All reasonable precautions have been taken by the World Health Organization to veriy the inormation contained in thispublication. However, the published material is being distributed without warranty o any kind, either expressed or implied. Teresponsibility or the interpretation and use o the material lies with the reader. In no event shall the World Health Organizationbe liable or damages arising rom its use.

Design and Layout : ww w.paprika-annecy.comPrinted in Geneva.



Contents

CForeword ......................................................................................................................................v

Abbreviations .............................................................................................................................vi

Acknowledgements................................................................................................................... vii

Glossary ......................................................................................................................................ix

1. Surveillance in dierent phases o malaria control ...........................................................1

1.1 Introduction ..............................................................................................................................1

1.2 Control phase: high- and moderate-transmission settings ................................................2

1.3 Control phase: low-transmission settings ............................................................................3

1.4 Elimination phase ....................................................................................................................4

2. Concepts o malaria surveillance in the control phase ......................................................6

2.1 Introduction ..............................................................................................................................6

2.2 Case denition ..........................................................................................................................7

2.3 Case detection...........................................................................................................................92.4 Surveillance indicators .......................................................................................................... 10

2.5 Limitations o surveillance data ........................................................................................... 14

2.6 Using surveillance data ......................................................................................................... 17

3. Data recording, reporting, analysis and use ....................................................................20

3.1 Recording ...............................................................................................................................20

3.2 Reporting.................................................................................................................................23

3.3 Data analysis ...........................................................................................................................24

3.4 Using data or making decisions .........................................................................................37

4. Establishing surveillance systems in the control phase ...................................................38

4.1 ools .........................................................................................................................................384.2 Procedures ...............................................................................................................................39

4.3 People .......................................................................................................................................39

4.4 Structures ................................................................................................................................39

Annexes .....................................................................................................................................43

Annex 1. ypes o malaria diagnostic test ..............................................................................45

Annex 2. Denition o severe malaria .....................................................................................48

Annex 3. Core surveil lance indicators or malaria control ..................................................49

Annex 4. Suggested register or community health workers, health postsand outpatient departments o health centres and hospitals ..................................55

Annex 5. Sheet or tallying outpatient attendance at health centres and hospitals ..........56Annex 6. Daily and weekly records o outpatient attendance at health centres

and hospitals ...............................................................................................................57

Annex 7. Discharge register or inpatient departments o health centresand hospitals ...............................................................................................................58

Annex 8. Reports rom health posts and community health workers tohealth acilities ...........................................................................................................59

Annex 9. Reports rom health acilities to the district level .................................................60

Annex 10. Line lists o inpatient malaria cases and deaths to be reported todistrict level in low-transmission settings..............................................................62

Annex 11. Line lists o all conrmed malaria cases to be reported at district level

in low-transmission settings ....................................................................................64Annex 12. Supervisory checklist or countries with high or moderate transmission ........66

Annex 13. Example quarterly bulletin or countries with high or moderate transmission ..67



Disease surveillanCe for malaria Control



foreworD

fd

Inormation on the number and distribution o malaria cases and deaths is critical or thedesign and implementation o malaria control programmes. It is needed to determine whichareas or population groups are most aected by malaria, so that resources can be targeted to thepopulations most in need. Inormation on the incidence o disease in relation to past levels isneeded to alert programmes about epidemics, so that control measures can be intensied. Dataon changes in disease incidence and mortality are also needed in order to judge the success o aprogramme and to determine whether it is perorming as expected or whether adjustments in the

scale or blend o interventions are required.

Te capacity o malaria surveillance systems to provide inormation on the distribution o andtrends in malaria varies widely across the globe. Te aim o this manual is to provide guidance tomalaria-endemic countries in designing and managing surveillance systems or malaria controland elimination, so that malaria programmes can obtain more complete, and more accurateinormation on malaria incidence, which can be used to help plan and monitor the programme.Te manual provides guidance on (i) the general principles that govern surveillance systems,including case denitions, procedures or case detection and investigation; (ii) data recording,reporting and analysis; and (iii) actors to be considered in establishing malaria surveillancesystems.

Recent developments in diagnostic testing present new opportunities or malaria surveillancesystems. Te availability o inexpensive, quality-assured rapid diagnostic tests or malariameans that parasite-based diagnosis is now possible not only at peripheral health-care acilitiesbut also at the community level. Tus, malaria surveillance can be based on conrmed ratherthan suspected cases at all levels o the health system. As malaria control measures expand andthe proportion o evers due to malaria alls rapidly, it becomes increasingly important to track conrmed malaria cases, rather than non-malarious evers, so that resources can be targeted toareas where problems remain and progress in malaria control is accelerated.

Te development o eective surveillance systems requires signicant investments, both nancialand human. A critical actor in the unctioning and sustainability o monitoring systems is theavailability o qualied, experienced personnel. Investment in data collection systems without

a commensurate investment in human resources to analyse the data and use the inormationgenerated is unlikely to yield signicant returns; ultimately, data should be used to inuencedecisions, and it is quality o the decisions rather than the quality o the data that will acceleratethe control o malaria.

Robert D. Newman, MD, MPH

Director, Global Malaria ProgrammeWorld Health Organization




abb

IN insecticide-treated net

LLIN long-lasting insecticidal net

PCR polymerase chain reaction

RD rapid diagnostic test

WHO World Health Organization



aCknowleDgements

acd

In 1994, a document rom the WHO Regional Oce or Arica described inormation systems orthe evaluation o malaria programmes and included indicators or inpatient (severe) malaria casesand malaria deaths.1 Te 20th report o the WHO Expert Committee on Malaria (1998) includedepidemiological indicators with standardized case denitions.2 Malaria was one o 40 diseasesincluded in the WHO recommended surveillance standards, published in 1999.3

In 2007–2008, WHO released three documents on malaria elimination that provided guidance onsurveillance.4,5,6 Inormation on surveillance during the elimination phase is also available romthe WHO Regional Oce or Europe (in Russian). WHO published recommended indicators ormonitoring malaria programmes in the World malaria report in 2008, 2009, 2010 and 2011.

In 2008, a document on disease surveillance and indicators or malaria control was prepared by the Global Malaria Programme at WHO. In April 2009, a WHO Global Malaria Programmeechnical Advisory Group reviewed the disease surveillance guidelines and indicators or thecontrol phase. Te group recommended that WHO also prepare surveillance and indicatorguidance or the elimination phase.

New versions o the manuals or disease surveillance or malaria control and elimination wereproduced in 2010 and revised in July 2011, beore being reviewed by WHO regional oces andexternal partners , including members o the Roll Back Malaria (RBM) Monitoring and Evaluation

Reerence Group (MERG).

Te ollowing WHO sta contributed to the production o the current manuals on malariasurveillance: Maru Aregawi, Richard Cibulskis, Charles Delacolette, Michael Lynch, RossitzaKurdova-Mintcheva, Mac Otten, Aae Rietveld and Ryan Williams.

Te WHO Global Malaria Programme echnical Advisory Group on Surveillance comprised:Abdul-Wahyd Ali (Zanzibar Malaria Control Programme); David Bell (Foundation or InnovativeNew Diagnostics); Elizabeth Chizema (Ministry o Health, Zambia); Erin Eckert (MacroInternational); Emmanuela Gakidou (Institute or Health Metrics and Evaluation); Pete Gethingand Anand Patil (University o Oxord); Korine Karema (National Malaria Control Programme,Rwanda); Eline Korenromp and Marcel Lama (Global Fund to Fight Aids uberculosis and

Malaria); Allan Schapira and om Smith (Swiss ropical and Public Health Institute); Steven

1 Inormation systems or the evaluation o malaria programmes. A practical guide. Brazzaville, WHO Regional Oceor Arica, 1994 (AFRO/CD/94.3). Available online at http://whqlibdoc.who.int/aro/1994-99/AFRO_CTD_

MAL_94.3.pd .2 Expert Committee on Malaria: twentieth report. Geneva, World Health Organization, 1998 (WHO echnical Report

Series No. 892). Available online at http://whqlibdoc.who.int/trs/WHO_TRS_892.pd .3 Available online at http://www.who.int/csr/resources/publications/surveillance/whocdscsrisr992.pd .4 Guidelines on the elimination o residual oci o malaria transmission . Cairo, WHO Regional Oce or the East-

ern Mediterranean, 2007 (EMRO echnical Publications Series 33). Available online at http://www.emro.who.int/

dsa/dsa742.pd .5 Malaria elimination: a eld manual or low and moderate endemic countries . Geneva, World Health Organization,

2007. Available online at http://whqlibdoc.who.int/publications/2007/9789241596084_eng.pd .

6 Global malaria control and elimination: report o a technical review. Geneva, World Health Organization, 2008.Available online at http://whqlibdoc.who.int/publications/2008/9789241596756_eng.pd .




Yoon (United States Centers or Disease Control and Prevention); Fiona Gore, Colin Mathers andRaman Velayudhan (WHO); Maru Aregawi, Richard Cibulskis, Mac Otten, Sergio Spinaci andRyan Williams (WHO Global Malaria Programme); Nathan Bakyaita (WHO Regional Oceor Arica); Rainier Escalada (WHO Regional Oce or the Americas); Ghasem Zamani (WHORegional Oce or the Eastern Mediterranean); Elkhan Gasimov (WHO Regional Oce or

Europe); and Charles Delacolette (WHO Regional Oce or South-East Asia).Documents were reviewed by the ollowing WHO sta: Andrea Bosman, Jo Lines, KaminiMendis, Abraham Mnzava, Sivakumaran Murugasampillay and Robert Newman (WHO GlobalMalaria Programme); George Ki-Zerbo, Nathan Bakyaita, Socé Fall and Etienne Minkoulo(WHO Regional Oce or Arica); Keith Carter and Rainier Escalada (WHO Regional Oceor the Americas); Hoda Atta and Ghasem Zamani (WHO Regional Oce or the EasternMediterranean); Mikhail Ejov and Elkhan Gasimov (WHO Regional Oce or Europe); LeonardOrtgea, Rakesh Rastogi and Kronthong Timasarn (WHO Regional Oce or South-EastAsia); Eva Christophel and Bayo Fatunmbi (WHO Regional Oce or the Western Pacic); andRabindra R. Abeyasinghe (WHO Country Oce Papua New Guinea).

Documents were also reviewed by: Andrei Baljaev (independent consultant), Marlize Coleman(Liverpool School o ropical Medicine), Erin Eckert (United States Agency or InternationalDevelopment), Scott Filler (Global Fund to Fight Aids uberculosis and Malaria), Roly Gosling(University o Caliornia), Ravi Goud (Macro International), Simon Hay (University o Oxord),Elatih Malik (Ministry o Health, Gezera State, Sudan), Steve Mellor (Malaria Consortium),Peter McIlroy and Steven Yoon (United States Centers or Disease Control and Prevention),Fabrizio Molteni (Research riangle International), Holly Newby (United Nations Children’sFund), Bruno Piotti and Deepika Kandula (Clinton Health Access Initiative), Allan Schapira(Swiss ropical Public Health), Rick Steketee (Malaria Control and Evaluation Partnership inArica) and Joshua Yukich (ulane University).

Funding or the production o this manual was grateully received rom the Government o Japan,

the United Kingdom Department or International Development, and the United States Agency or International Development.



glossary

g

Active case detection: Te detection by health workers o malaria inections at community andhousehold level in population groups that are considered to be at high risk. Active case detectioncan be conducted as ever screening ollowed by parasitological examination o all ebrile patientsor as parasitological examination o the target population without prior ever screening.

Annual blood examination rate: Te number o examinations o blood slides or malaria by microscopy per 100 population per year.

Case-based surveillance: Every case is reported and investigated immediately (and also includedin the weekly reporting system).

Case denition (control programmes)

Conrmed malaria: Suspected malaria case in which malaria parasites have beendemonstrated in a patient’s blood by microscopy or a rapid diagnostic test.

Presumed malaria: Suspected malaria case without a diagnostic test to conrm malaria butnevertheless treated presumptively as malaria.

Suspected malaria: Patient illness suspected by a health worker to be due to malaria. Tecriteria usually includes ever. All patients with suspected malaria should receive a diagnostictest or malaria, by microscopy or a rapid diagnostic test.

Case denition (elimination programmes)

Autochthonous: A case acquired by local transmission, i.e. an indigenous or introduced case(also called ‘ locally transmitted’).

Imported: A case the origin o which can be traced to a known malarious area outside thecountry in which the case was diagnosed.

Indigenous: Any case contracted locally (i.e. within national boundaries), without strongevidence o a direct link to an imported case. Tese include delayed rst attacks o P. vivaxmalaria due to locally acquired parasites with a long incubation period.

Induced: A case the origin o which can be traced to a blood transusion or other orm o parenteral inoculation but not to normal transmission by a mosquito.

Introduced: A case contracted locally, with strong epidemiological evidence linking itdirectly to a known imported case (rst generation rom an imported case, i.e. the mosquitowas inected rom a case classied as imported).

Locally transmitted: A case acquired by local transmission, i.e. an indigenous or introducedcase (also called ‘autochthonous’).

Malaria: Any case in which, regardless o the presence or absence o clinical symptoms,malaria parasites have been conrmed by quality-controlled laboratory diagnosis.

Case investigation: Collection o inormation to allow classication o a malaria case by origin

o inection, i.e. whether it was imported, introduced, indigenous or induced. Case investigationincludes administration o a standardized questionnaire to a person in whom a malaria inectionis diagnosed.




Case management: Diagnosis, treatment, clinical care and ollow-up o malaria cases.

Case notication: Compulsory reporting o detected cases o malaria by all medical units andmedical practitioners, to either the health department or the malaria elimination service (as laiddown by law or regulation).

Certication o malaria-ree status: Granted by WHO afer proo beyond reasonable doubtthat the chain o local human malaria transmission by Anopheles mosquitoes has been ully interrupted in an entire country or at least 3 consecutive years.

Control charts: Figures summarizing inormation on key malaria indicators collected by surveillance or regular, periodic review by malaria control programme personnel.

Discharge register: List o patients who leave inpatient hospital care. Discharge registers shouldcontain the date o admission, patient’s name, residence, age, sex, diagnosis, length o stay and reason or leaving (discharged, died, transerred, absconded). Tis inormation should beabstracted rom the patient le by appropriately trained sta.

Elimination: Reduction to zero o the incidence o inection by human malaria parasites in

a dened geographical area as a result o deliberate eorts. Continued measures to preventre-establishment o transmission are required.

Endemic: Applied to malaria when there is an ongoing, measurable incidence o cases andmosquito-borne transmission in an area over a succession o years.

Epidemic: Occurrence o cases in excess o the number expected in a given place and time.

Eradication: Permanent reduction to zero o the worldwide incidence o inection caused by human malaria parasites as a result o deliberate eorts. Intervention measures are no longerneeded once eradication has been achieved.

Evaluation: Attempts to determine as systematically and objectively as possible the relevance,eectiveness and impact o activities in relation to their objectives.

False negative (or alse positive): A negative (or positive) result in a test when the opposite is true.

Focus: A dened, circumscribed locality situated in a currently or ormer malarious areacontaining the continuous or intermittent epidemiological actors necessary or malariatransmission. Foci can be classied as endemic, residual active, residual non-active, cleared up,new potential, new active or pseudo.

Gametocyte: Te sexual reproductive stage o the malaria parasite present in the host’s red bloodcells.

Incubation period: Te time between inection (by inoculation or otherwise) and the rstappearance o clinical signs, o which ever is the commonest.

Intervention (public health): Activity undertaken to prevent or reduce the occurrence o a healthcondition in a population. Example o interventions or malaria control include the distributiono insecticide-treated mosquito nets, indoor residual spraying with insecticides, provision o eective antimalarial therapy or prevention or curative treatment o clinical malaria.

Line list: Inormation on cases recorded in rows and columns, with data or each case in columnsacross one row. Te inormation may include case identication number; demographic actors(patient’s name, address, age, sex); clinical actors (date o attendance, type o test, test result,treatment received); intervention actors (house sprayed, insecticide-treated net ownership,preventive therapy).

Local mosquito-borne malaria transmission: Occurrence o human malaria cases acquired in a

given area through the bite o inected Anopheles mosquitoes.



glossaryglossary

Malaria-ree: An area in which there is no continuing local mosquito-borne malaria transmission,and the risk or acquiring malaria is limited to introduced cases.

Malaria incidence: Te number o newly diagnosed malaria cases during a specied time in aspecied population.

Malaria prevalence: Te number o malaria cases at any given time in a specied population,measured as positive laboratory test results.

Monitoring (o programmes): Periodic review o the implementation o an activity, seeking toensure that inputs, deliveries, work schedules, targeted outputs and other required actions areproceeding according to plan.

National ocus register: Centralized computerized database o all malaria oci in a country.

National malaria case register: Centralized computerized database o all malaria cases registeredin a country, irrespective o where and how they were diagnosed and treated.

Outpatient register: List o patients seen in consultation in a health acility, which may includethe date o the consultation, patient’s age, place o residence, presenting health complaint, test

perormed and diagnosis.

Parasite prevalence: Proportion o the population in whom Plasmodium inection is detected ata particular time with a diagnostic test (usually microscopy or a rapid diagnostic test).

Passive case detection: Detection o malaria cases among patients who on their own initiativewent to a health post or treatment, usually or ebrile disease.

Population at risk: Population living in a geographical area in which locally acquired malariacases occurred in the current and/or previous years.

Rapid diagnostic test: An antigen-based stick, cassette or card test or malaria in which a colouredline indicates that plasmodial antigens have been detected.

Rapid diagnostic test positivity rate: Proportion o positive results in rapid diagnostic testsamong all the tests perormed.

Receptivity: Sucient presence o anopheline vectors and existence o other ecological andclimatic actors avouring malaria transmission.

Re-establishment o transmission: Renewed presence o a constant measurable incidence o cases and mosquito-borne transmission in an area over a succession o years. An indication o thepossible re-establishment o transmission would be the occurrence o three or more introducedand/or indigenous malaria inections in the same geographical ocus, or 2 consecutive years orP. alciparum and or 3 consecutive years or P. vivax.

Relapse (clinical): Renewed maniestation o an inection afer temporary latency, arising romactivation o hypnozoites; thereore limited to inections with P. vivax and P. ovale.

Sensitivity (o a test): Proportion o people with malaria inection (true positives) who have apositive test result.

Slide positivity rate: Proportion o slides ound positive among the slides examined.

Specicity (o a test): Proportion o people without malaria inection (true negatives) who havea negative test result.

Surveillance (control programmes): Ongoing, systematic collection, analysis and interpretationo disease-specic data or use in planning, implementing and evaluating public health practice.




Surveillance (elimination programmes): Tat part o the programme designed or theidentication, investigation and elimination o continuing transmission, the prevention and cureo inections and nal substantiation o claimed elimination.

Transmission intensity: Rate at which people in a given area are inoculated with malaria parasitesby mosquitoes. Tis is ofen expressed as the ‘annual entomological inoculation rate’, which is the

number o inoculations with malaria parasites received by one person in 1 year.

Transmission season: Period o the year during which mosquito-borne transmission o malariainection usually takes place.

Vector control: Measures o any kind against malaria-transmitting mosquitoes intended to limittheir ability to transmit the disease.

Vector efciency: Ability o a mosquito species, in comparison with another species in a similarclimatic environment, to transmit malaria in nature.

Vectorial capacity: Number o new inections that the population o a given vector wouldinduce per case per day at a given place and time, assuming conditions o non-immunity. Factors

aecting vectorial capacity include: (i) the density o emale anophelines relative to humans; (ii)their longevity, requency o eeding and propensity to bite humans; and (iii) the length o theextrinsic cycle o the parasite.

Vigilance: A unction o the public health service during a programme or prevention o re-introduction o transmission, consisting o watchulness or any occurrence o malaria in anarea in which it had not existed, or rom which it had been eliminated, and application o thenecessary measures against it.

Vulnerability: Either proximity to a malarious area or requent inux o inected individuals orgroups and/or inective anophelines.



1surveillanCe in Different phases of malaria Control

1. sc d c

1.1 Introduction

A malaria surveillance system consists o the tools, procedures, people and structures thatgenerate inormation on malaria cases and deaths, which can be used or planning, monitoringand evaluating malaria control programmes. An eective malaria surveillance system enablesprogramme managers to:

• identiy the areas or population groups most aected by malaria;

• identiy trends in cases and deaths that require additional intervention, e.g. epidemics; and

•assess the impact o control measures.

With this inormation, programmes can direct resources to the populations most in need andrespond to unusual trends, such as outbreaks o cases or the absence o a decrease in the numbero cases despite widespread implementation o interventions. As a result, progress in malariacontrol can be accelerated and wastage o resources avoided.

Te design o malaria surveillance systems depends on two actors: (i) the level o malariatransmission and (ii) the resources available to conduct surveillance. In the initial phase o control, there are ofen so many malaria cases that it is not possible to examine and react to each

conrmed case individually: rather, analysis is based on aggregate numbers, and action is taken ata population level. As transmission is progressively reduced, it becomes increasingly possible, andnecessary, to track and respond to individual cases. able 1 illustrates the way in which malariasurveillance is undertaken in dierent transmission settings and phases o control.

Table 1 Malaria surveillance in different transmission settings and phases of control

Control phase Elimination phase

Transmission: High & moderate Low Very low

Parasite prevalence(2-9 yrs):

>10% <10%

Incidence: C d d c dccd <5

C d d cd dbd ccd d

C dc

ld vb d b

r dc

id c

ld c gc

Ccd

fc db

Fevers: p d

p d

p d

Health facilityattendance:

h d l d

Vectors: ec Cd c/ c Cd c/ c

Aims of program: m & c dc C dc e

1



2 Disease surveillanCe for malaria Control

Table 1 Malaria surveillance in different transmission settings and phases of control

Surveillancesystem:

Resources: l d d

l q d ccb c

wdd b dc d

rc cc

Data recording: a b a b

l d d

→ c

C d

Investigation: i c i c→ c idd c

Te term ‘high transmission’ has usually been used to indicate hyper- and holoendemicmalaria (parasite prevalence in children aged 2–9 years > 50%), ‘moderate transmission’ toindicate mesoendemic malaria (10–50% parasite prevalence) and ‘low transmission’ to indicatehypoendemic malaria (parasite prevalence < 10%)1,2. Te threshold o 10% is used to characterizelow transmission in this manual or consistency and to provide a general guide to the types o malaria surveillance possible at dierent levels o malaria endemicity. Te thresholds are not,however, xed, and surveillance strategies or low-transmission settings might sometimes bemore appropriately undertaken when parasite prevalence is < 5% rather than < 10%.

1.2 Control phase: high- and moderate-transmission settings

High- and moderate-transmission settings are generally characterized by: (i) a concentration o malaria cases and deaths in children under 5 years o age, with pregnant women also susceptibleto the eects o malaria; (ii) a high proportion o cases due to Plasmodium alciparum; (iii)suspected malaria comprising a high percentage (typically > 20%) o outpatient attendances,hospital inpatients and recorded deaths; (iv) a high proportion o evers due to malaria, (althoughgenerally < 30%); (v) high parasite prevalence rates in children (> 10%); (vi) ecient anopheline

vectors; and (vii) a high requency o malaria-related deaths.

High- and moderate-transmission settings are ofen ound in low-income countries, which havelow expenditures per person on health care services. Tis results in weak health systems thatare not easily accessed by the population, lower sta to patient ratios, requent interruptionso medical supplies and limited use o parasitological diagnosis. In such settings, the primary emphasis o malaria programmes has ofen been on reducing mortality (by prevention andappropriate management o severe cases) and the secondary emphasis on case reduction. Teeatures o surveillance in high-transmission settings are shown in Box 1.1.

1 Terminology o malaria and o malaria eradication. Report o a draing committee. Geneva, World HealthOrganization, 1963.

2 Hay SI, Smith DL, Snow RW. Measuring malaria endemicity rom intense to interrupted transmission. Lancet Inectious Diseases, 2008, 8:369–378.




Box 1.1.

Features of malaria surveillance systems in the control phase: high- and moderate-transmission settings

Registers o individual cases are maintained at health acilities, which allow recording o

diagnostic tests perormed and test results. Given the high requency o malaria cases andthe limited resources or maintaining an extensive recording and reporting system, malariasurveillance systems rely on the reporting and use o aggregate data by district and higheradministrative levels. Malaria surveillance is requently integrated into a broader system o health inormation or communicable disease surveillance.

At the health acility level, case-based surveillance o malaria inpatient cases and deathsis undertaken with the aim o responding to cases o severe disease and attaining a targeto zero malaria deaths. Cases are graphed monthly to assess the extent to which controlmeasures are reducing the incidence o malaria.

At district and national levels, cases and deaths are summarized monthly on ve control

charts, in order to assess the ecacy o malaria control interventions and identiy trendsthat require an urgent response. Te control charts cover: (i) malaria incidence and mortality rates; (ii) proportional malaria incidence and mortality rates; (iii) general patient attendancerates; (iv) diagnostic activity (annual blood examination rate); and (v) quality o diagnosisand health acility reporting. Analysis is also undertaken by health acility catchment areaand by district in order to set priorities or malaria control activities.

1.3 Control phase: low-transmission settings

Low-transmission settings are characterized by (i) a lower incidence o conrmed malaria cases;(ii) a more uniorm spread o cases by age or more concentrated in population groups withhigher exposure; (iii) lower malaria mortality rates; (iv) a parasite prevalence in children aged2–9 years < 10%; (v) generally more seasonal malaria, with a higher risk o epidemics;1 (vi) asmall proportion o evers attributable to malaria, especially in the low-transmission season; (vii)malaria distribution more ocal within districts; and (viii) imported cases comprising a signicantproportion o all cases. In some temperate and subtropical areas, P. vivax may occur in higherproportions, particularly as P. alciparum disappears more quickly than P. vivax in response tocontrol measures.

Health systems in low-transmission settings are usually stronger than in high-transmissionsettings, and there may be widespread availability o parasitological diagnosis and appropriate

treatment. Malaria may, however, be concentrated in marginalized populations, such as thoseliving in remote border areas, migrant workers and tribal populations, and innovative ways may have to be ound to reach these groups. Te eatures o malaria surveillance in low-transmissionsettings are shown in Box 1.2.

1 Te two types o setting with low transmission are (i) locations in t ransition rom high or moderate transmission tolow transmission and (ii) locations that have had low transmission or many years because o environmental actors

or less ecient vectors. Epidemics may be more likely in areas in which malaria control has been successul and inwhich ecient anopheline vectors remain.

1




Box 1.2.

Features of malaria surveillance systems in the control phase: low transmissionsettings

Registers o individual malaria cases are maintained at health acilities, with records o

the diagnostic tests perormed and the results. As well as aggregate data being reportedto district and higher administrative levels, line lists o inpatients and inpatient deaths areorwarded to district level, and, when case loads and district capacity permit (or example,< 150 patients per district per month), lists o all conrmed cases are submitted monthly.

At health acility level, case-based surveillance o malaria cases and deaths is undertaken,with the aim o identiying population groups with the highest malaria incidence andprobable sources o inection. Cases are graphed daily or weekly to identiy trends thatrequire attention and are mapped by village to identiy clusters o cases.

At the district level, malaria cases and deaths are summarized weekly or monthly on thesame ve control charts used in high-transmission settings, in order to assess the impact o

malaria control interventions and identiy trends that require urgent response. Te controlcharts cover: (i) malaria incidence and mortality rates; (ii) proportional malaria incidenceand mortality rates; (iii) general patient attendance rates; (iv) diagnostic activity (annualblood examination rate); and (v) quality o diagnosis and health acility reporting. Analysisis undertaken by health acility catchment area and by village in order to set priorities oractivities. A register o severe cases and deaths is maintained and investigations undertakento identiy and address programme weaknesses.

At national level, cases and deaths are summarized monthly on the ve control chartsin order to assess the impact o malaria control interventions. Analysis is undertaken by district in order to set priorities or activities.

1.4 Elimination phase

In the elimination phase, cases occur sporadically or in distinct oci. Imported cases may comprisea signicant proportion o all cases and may pose a risk or re-establishment o transmission inareas in which it had previously been interrupted. Countries have resources to investigate eachcase to ascertain whether it is imported or locally acquired and undertake appropriate controlmeasures. Box 1.3 shows the eatures o surveillance in elimination settings.




Box 1.3.

Features of malaria surveillance systems in the elimination phase

Case-based surveillance is perormed. Each conrmed case is immediately notied to district,provincial and central levels. A ull investigation o each case is undertaken to determine

whether it was imported, acquired locally by mosquito-borne transmission (introduced,indigenous, relapsed) or induced. Te national reerence laboratory reconrms all positivetest results and a sample o negative test results and organizes laboratory participation in anational quality assurance network.

Each new ocus o transmission is investigated, including an entomological investigation,to ascertain risk actors and devise the optimal strategies or control. Te ocus is classied,and its status is updated continuously.

Te malaria programme monitors the extent o surveillance, mainly by tracking bloodexamination rates by village and by month in high-risk oci and comparing the number o diagnostic tests done with the number expected.

Programme managers at district level keep: (i) malaria case investigation orms, patientrecords, ocus investigation orms and a register o oci with changes in status; (ii) mapsshowing the distribution o cases by household, vector breeding places, possible sites o transmission and geographical eatures, such as hills, rivers and roads; and (iii) data onintegrated vector control interventions.

Full documentation o programme activities and surveillance results is kept securely atnational level in preparation or certication o malaria elimination.

Tere are no strict rules about when countries change their approach to surveillance. Tisdepends on the level o malaria transmission and the capacity o the control programme toperorm specic surveillance activities. Some countries in relatively high-transmission settingsmay adopt certain approaches used in low-transmission settings, and their control programmeswould be expected to progress more rapidly as a result o better targeting o interventions. Many low-transmission countries may wish to adopt certain approaches used in the elimination phase.Dierent approaches may be used in dierent settings within a country, particularly whentransmission intensity varies geographically.

Tis manual describes the general principles that govern surveillance systems in the controlphase (in high-, moderate- and low-transmission settings), including case denitions, proceduresor case detection and indicators or programme management. It also outlines recommended

practices or recording, reporting and analysing data and presents actors to be considered whenestablishing surveillance systems in the control phase. Te companion manual, entitled Diseasesurveillance or malaria elimination, covers similar topics in settings in which the programme isoriented towards eventual malaria elimination.

1




2. Cc c c

2.1 Introduction

Te objective o malaria control is to reduce the incidence o and mortality rom malaria asrapidly and economically as possible. Surveillance systems can help programme managers to dothis, by providing inormation on the populations in which the incidence o malaria is highest(and thereore to whom resources should be targeted) and on changes in incidence over time thatrequire attention.

Te main source o inormation or malaria surveillance in the control phase is reports o conrmed malaria cases, malaria inpatients and malaria deaths obtained rom all or selectedpublic sector health acilities. Tese may be complemented by data rom household surveys on theprevalence o parasitaemia and intervention coverage.

In high- and moderate-transmission areas, monthly counts o malaria cases, inpatients anddeaths can be used to determine trends over time and the geographical distribution o malaria.At health acility level, data on individual patients are used to investigate the circumstancessurrounding each admitted case and death, so as to identiy programme weaknesses andpotential improvements. As transmission is reduced and the risk o epidemics increases, morerequent analysis o cases is undertaken at health acility level to allow early detection o potentialoutbreaks. Moreover, as the numbers o severe cases and deaths diminish, health acilities canreport details o each malaria inpatient and death to district level so that a district register o severe cases can be assembled and action taken to address persistent problems.

In low-transmission areas when there is appreciable heterogeneity in the distribution o malaria,it becomes increasingly important to identiy the population groups most susceptible to inectionand to target resources appropriately. When the case incidence is reduced suciently, healthacilities can begin to report details o individual malaria cases to district level. Tese reportscan be used to construct a case register that provides more detailed inormation on the principallocations and population groups aected by malaria.

In the initial phases o building an eective malaria surveillance system, attention will ocus onensuring good-quality data. Tis involves making sure that all people with suspected malaria

receive a diagnostic test, that cases are correctly classied according to the test result, that thereis a quality management system or both microscopy and rapid diagnostic tests (RDs), andthat registration and reporting rom health acilities are complete and consistent. Te quality o surveillance systems must be monitored continuously by maintaining an up-to-date list o operational health acilities, keeping track o which acilities have submitted the required reports,ollowing up on missing reports, reviewing the data submitted and ollowing up on incompleteor erroneous data as well as providing positive eedback to health acilities that submit timely,complete, accurate data. Data rom surveillance must also be interpreted careully to identiy any weaknesses in systems.



7ConCepts of malaria surveillanCe in the Control phase

2.2 Case definition

Malaria cases

Although WHO now recommends that all suspected cases o malaria be conrmed with a diagnostictest beore treatment (the dierent types o diagnostic test are described in Annex 1), this is not yet

the practice in all settings, either because access to diagnostic testing is not yet available or becauseo stock-outs o RDs or the materials necessary to prepare and examine blood lms by microscopy.Tus, it is necessary to distinguish between suspected malaria cases, presumed cases and conrmedcases. Te relations among these categories are shown diagrammatically in Figure 1.

figure 1.

Suspected

TestedNot tested(presumed)

Positive(confirmed)

Negative(not malaria)

Suspected malaria case: In a suspected case, a patient is suspected by a health worker o havingmalaria. Te criteria or suspected malaria usually include ever or a history o ever, but theprecise criteria vary according to local circumstances and are established by the national malariacontrol programme.1 All suspected cases o malaria are tested by either microscopy or an RD.

Presumed (not tested) malaria case: In a suspected malaria case, the patient did not receive adiagnostic test or malaria but was nevertheless treated or malaria. Such cases have also beenreerred to as ‘probable’ cases;2 however, in most settings, the chance that a suspected case will beconrmed is < 50%, and thereore use o the term ‘probable’ is inappropriate. Such cases are alsosometimes reerred as ‘unconrmed’ cases. In this guide, the term used is ‘presumed malaria case’.

Conrmed malaria case: A suspected case o malaria in which malaria parasites have beendemonstrated, generally by microscopy or a RD, becomes a conrmed case. Te denition impliesthat the patient displayed symptoms o malaria, and the presence o parasites was conrmed. Insome suspected cases with a positive test, particularly in populations that have acquired immunity to malaria, ebrile illness may be due to other causes. Nevertheless, a diagnosis o conrmed malaria

is still given. I a concurrent disease is suspected, it should be urther investigated and treated.Not malaria (conrmed not to be malaria): Patients with suspected malaria or whom a diagnostic testwas negative would usually be given a diagnosis other than malaria. It is possible that some patientswho test negative by microscopy or RD have very low levels o parasitaemia that are detectable only by more sensitive techniques, such as polymerase chain reaction (PCR)3 testing. Microscopy or RDmight have to be repeated i no other source o ever is identied and the symptoms continue. Suchlow levels o parasitaemia are generally considered not to be clinically signicant in most settings, anddiagnostic testing with microscopy or RD should allow adequate tracking o malaria trends.

1 Universal access to malaria diagnostic testing; an operational manual . Geneva, World Health Organization, 2011.2 WHO expert committee on malaria: twentieth report . Geneva, World Health Organization, 1998.

3 Polymerase chain reaction is a highly sensitive test or detecting very small amounts o genetic material romparasites.

2




As a high proportion o suspected and presumed cases (generally > 70%1) are not malariousevers, these counts do not provide good measures or malaria surveillance (see Box 2.1). Malariasurveillance should thereore be based on conrmed cases. It is also important to report thedierent categories (suspected, presumed and conrmed) separately; it is not helpul to add thesenumbers (e.g. to report presumed plus conrmed cases), as the nal values are not comparable

over time when the incidence o malaria in the community changes.

Box 2.1.

Advantage of focusing on confirmed cases of malaria

TestedSuspected cases

decreased by 25%

Confirmed negative

Confirmed malaria

Confirmed cases

decreased by 50%

Suspected

malaria cases

2010 2011 2012 2013 2014

250

200

150

100

50

0

Te example above shows trends in malaria in a district in which 100% o suspected caseswere tested or malaria. It can be seen that non-malarial evers comprised 50% o allsuspected malaria cases in 2010. Between 2010 and 2014, it is projected that the number o conrmed malaria cases will drop by 50%, while that o non-malarial evers will remainconstant. Te number o suspected cases will drop by only 25%, and most will not be due tomalaria. Tus, surveillance o conrmed malaria cases is ar more sensitive to programmechange than surveillance o suspected cases.

Not Tested

Tested

Suspectedmalaria cases

2010 2011 2012 2013 2014

250

200

150

100

50

0

Presumed malaria (not tested)

Confrimed negative

Confirmed malaria

Presumed casesdecreased by 25%

Confirmed cases

decreased by 50%

Te example above shows trends in malaria or a district with the same trends in malariaas above but in which testing is done on only 20% o suspected cases o malaria. Althoughthe numbers o conrmed cases detected will be smaller, the trend is similar to that above:between 2010 and 2014, the number o conrmed malaria cases is predicted to drop by 50%;however, the number o presumed or untested cases will drop by only 25% and, i tested,most would not be malaria in 2014. Similarly, the number o presumed plus conrmed caseswill drop by only 25%. Tus, surveillance o conrmed malaria cases is more sensitive toprogramme change than surveillance o presumed cases.

1 D’Acremont V, Lengeler C, Genton B. Reduction in the proportion o evers associated with Plasmodium alciparum parasitaemia in Arica: a systematic review. Malaria Journal , 2010, 9:240.




Severe malaria cases and deaths

Malaria cases can be categorized as uncomplicated or severe. Te clinical eatures o severe malariaare listed in Annex 2. In general, people with uncomplicated malaria are treated as outpatients,while those with severe malaria are managed as inpatients.1 For surveillance purposes duringthe control phase, thereore, outpatient and inpatient malaria cases are considered proxies or

uncomplicated and severe malaria, respectively.2

Te numbers o inpatient malaria cases and deaths should be taken rom the register o dischargesin which malaria is the primary diagnosis. I a discharge register is not kept, inpatient casesmay be recorded rom ward books, although the diagnosis at the time o admission should notbe used or surveillance, as this may be presumed rather than conrmed malaria. Inpatientswith a primary discharge diagnosis o malaria should have had a positive test or malaria duringhospitalization. I parasite-based testing is not available, the discharge diagnosis on clinicalgrounds and response to treatment are used to assign the discharge diagnosis. Te predictive valueo a discharge diagnosis on clinical grounds or severe malaria is considered to be higher thanor uncomplicated malaria and can be used or surveillance purposes i testing is not available.Diagnostic testing should be introduced or inpatients as a priority, to enhance not only malaria

surveillance but also treatment outcomes.

Te numbers o inpatients and deaths at all hospitals and health centres with beds should bereported.

2.3 Case detection

Cases can be detected passively or actively.

Passive case detection

Passive case detection is the regular or periodic collection o data rom case reports or registers in

health care acilities at which patients seek care at their discretion. Passive case detection can alsoinclude mobile health services at dened posts, additional xed health posts in high-transmissionor problem areas and treatment in community-based programmes at which patients seek care attheir discretion.

Active case detection

Active case detection involves searching or malaria cases and diagnostic testing at the community or household level by health workers on regular or occasional visits. esting may be conned topatients with ever, or everyone may be tested (mass screening). Active case detection can bedone to ll gaps in passive case detection systems (e.g. to detect cases in populations with limitedaccess to services, such as migrant populations). Tis is sometimes known as ‘proactive’ case

detection, in which a population is examined even though there may be no evidence o conrmedcases. Active case detection may also be undertaken in response to a conrmed case or clustero cases, in which a dened population potentially linked to a conrmed case is identied, andsymptomatic cases are tested (possibly with a RD then by blood slide or conrmation) as well asasymptomatic cases (by blood slides only). Tis is sometimes known as ‘reactive’ case detection.

1 Some countries with low transmission and in the elimination phase might admit uncomplicated malaria cases toensure ull adherence to treatment.

2 Te use o inpatient malaria cases as a proxy or severe malaria in Arica is described in Inormation systems or theevaluation o malaria programmes. A practical guide. Brazzaville, WHO Regional Oce or Arica, 1994 (AFRO/

CD/94.3).

2




In general, malaria surveillance in the control phase relies on passive case detection, particularly monthly or weekly reports o conrmed malaria cases submitted by public health acilities. Tesemay be supplemented by reports rom private health acilities run by religious organizations,mining companies or other organizations that have an agreement to report to the ministry o health. Reports o cases tested and treated by community health workers should also be included

in the surveillance system, when such programmes exist.Monthly reports rom health acilities should distinguish between cases detected passively,actively and in the community; otherwise, trends in the number o cases could be aected by theextent o active case detection undertaken each month or irregular reporting rom community health workers.

As malaria incidence decreases and becomes increasingly ocused in marginalized populations,active case detection and community health workers are likely to have more important roles inidentiying cases.

2.4 Surveillance indicators

In the initial phase o malaria control in high-transmission settings, there are so many malariacases that it is not possible to examine and react to each case individually; rather, much analysisis based on aggregate counts o cases and deaths, and action is taken at a population level, e.g.deciding which populations would benet rom additional measures, such as indoor residualspraying.1 Te counts might, however, have to be adjusted to take into account population size,diagnostic activity or other actors, thus transorming numbers into ‘indicators’, so that they provide more meaningul inormation (see Box 2.2). en indicators are particularly useul ormalaria surveillance in the control phase (see Annex 3 or ull denitions):

1. Number of confirmed malaria cases per 1000 population per month or per year

Tis indicator can be calculated by month and by year. Te number o malaria cases uctuateswith the transmission season; it can be useul in assessing the success o preventive programmesand demand or treatment in the public sector. Te variable is, however, sensitive to changes inreporting rates, diagnostic practice and use o health acilities. Care should be taken to ensurethat reporting has been consistent over time, by examining trends in health acility reportingrates, annual blood examination rates and total outpatient attendance. I these indicators havechanged, it may be more inormative to examine trends in test positivity rates (slide or RD) orconne the analysis to a subset o health acilities that have reported consistently over time.

1 An exception would be the investigation o severe cases and deaths, which would initially be undertaken at healthacility level with support rom districts.




Box 2.2.

Adjusting for population size: calculating incidence rates

Absolute numbers o malaria cases, inpatients and deaths can be used to examine trends overtime and can help to identiy places in which the problem o malaria is greatest. Absolute

numbers are less useul or assessing which populations are at the highest risk or acquiringmalaria, as most geographical units have dierent population sizes. For example, it is dicultto decide whether 100 cases in a population o 3400 represents a higher risk or acquiringmalaria than 270 cases in a population o 8500. In order to make comparisons betweenpopulations easily, the number o cases is usually expressed or a standard population o 1000 or 10 000, by dividing the number o cases by the population size and multiplying by the standard size o population desired.

Population A: 100 cases per 3400 population x 1000 = 29.4 cases per 1000 population

Population B: 270 cases per 8500 population x 1000 = 31.8 cases per 1000 population

Adjustment to a standard population can also be used to take into account the growth o populations over time, which may be signicant when examining trends in cases over anextended period such as 10 years.

Te denominator is generally the population at risk or malaria. Tis is dened as thepopulation in areas in which there is ongoing transmission o malaria. People travellingto such areas may acquire malaria, but these are not normally included in the populationat risk. For international comparisons and other situations in which inormation on theoverall risk to populations is desired (including those not exposed to malaria), the total population o a country may be used as the denominator. I cases are broken down by age,sex or occupational group, the size o these groups should be used as the denominator.

Programme managers may also be interested in knowing the size o other populations (e.g.those living in areas where vectors are circulating or target populations or interventions),but these are generally not used or calculating incidence rates.

Estimates o population size published by the relevant government department should beused (e.g. the statistical oce, planning bureau or census oce). Tese are usually based onprojections rom censuses undertaken at intervals o approximately 10 years. Populationgrowth rates between censuses are used to project population sizes afer the latest census.Tus, as the time or the next census approaches, the population projections may deviateconsiderably rom the actual population sizes, particularly at local level. When new censusresults are released, the projected populations calculated or previous years must be updatedto take into account the latest—and more accurate—counts.

2




2. Number of inpatient malaria cases per 10 000 population per month or per year and

3. Number of inpatient malaria deaths per 100 000 population per month or per year

Tese two variables are indicators o severe malaria and deaths and are thereore also importantor judging the success o malaria programmes. In order to monitor trends, it is ofen moreinormative to examine the number o inpatient cases, owing to the comparative rarity o deaths. For instance, a district with a population o 100 000 and a crude birth rate o 35 per 1000population, a rate o mortality o children under 5 years o 100 per 1000 births and 25% o deathsdue to malaria may expect ewer than 9 malaria deaths per year. I 20% o the deaths occur inhealth acilities, a district will expect to record ewer than 20 malaria deaths per year (or two permonth), although they may be clustered according to seasons with the highest transmission. Incontrast, with a case atality rate o 2%, the same district could see 100 inpatient malaria casesper month.

Te numbers o inpatient malaria cases are known to uctuate by malaria transmission seasonand are sensitive to changes due to malaria control activities, decreasing rapidly when highcoverage with interventions has been achieved.1 Te rapidity o changes in the numbers o malaria

inpatients and deaths might also be inuenced by the initial level o transmission, with moregradual change in areas with the highest transmission intensity and higher parasite prevalencein children.

Inpatient cases should be conrmed by parasitological diagnosis. In situations where parasitologicaltesting is not common, an inpatient diagnosis o malaria is nevertheless considered to be morespecic than an outpatient diagnosis, and, despite the possibility o overdiagnosis, trends ininpatient malaria cases are likely to reect real changes.

As or malaria cases, care should be taken to ensure that reporting o inpatient cases has beenconsistent over time. It is thereore important to examine trends in health acility reporting rates,as well as total numbers o inpatients and deaths. I there have been changes in these indicators,

it may be more inormative to examine trends in the proportions o inpatients and deaths due tomalaria or to conne the analysis to the subset o health acilities that have reported consistently over time.

rends in the numbers o inpatient malaria cases and deaths should align with the total numbero conrmed malaria cases, and any dierences should be investigated. Tese may be due to:reporting issues (e.g. some health acilities may not report inpatient data), dierences in diagnosticpractice over time or real changes (e.g. a decrease in the proportion o cases becoming severebecause o improved access to more timely, eective treatment).

4. Malaria test positivity rate (RDT and/or slide positivity rate)

Tis indicator can provide inormation on trends in malaria. In some settings, slide positivity

rates have decreased rom 30–60% to < 10% in response to control measures implemented in theprevious 2–3 years. est positivity rates can vary by season, and the peak test positivity rate seenduring a year might be quite dierent rom the annual average.

Both slide and RD positivity rates are less sensitive to changes in reporting rates, diagnosticpractices and health acility utilization rates than trends in conrmed cases or incidence rates(because data that are changing are excluded rom both the numerator and denominator). Forthis reason, they may be more helpul in identiying areas in which malaria transmission is mostintense than malaria incidence rates (which are particularly aected by the accessibility anduse o health acilities as well as reporting rates). Tey are not, however, immune to distortion.For example, test positivity rates can increase i parasitological diagnosis has been extended to

1 World malaria report 2011, Geneva, World Helath Organization, 2011.




populations living in more intense transmission areas where testing was not available previously.Attention should also be paid to the quality o diagnostic testing and potential changes overtime; in some health acilities, poor-quality microscopy can lead to considerable overdiagnosis o malaria.1 Tereore, possible errors and conounding actors should be taken into account wheninterpreting trends.

5. Percentage of cases due to P. falciparum

In areas in which more than one species o Plasmodium is present, it is useul to monitor thepercentage o cases due to P. alciparum, as this can provide inormation on the extent o malariacontrol, the likelihood o observing severe cases and the extent to which the programme shouldbe adjusted to address P. vivax or other species. In areas where control measures are intensied,the proportion o cases due to P. alciparum may decrease; P. vivax appears to be respond lessquickly to control measures because it can tolerate a wider range o environmental conditionsand because the dormant liver stage (hypnozoite) enables inections to persist in the absence o mosquito transmission. P. ovale and P. malariae may also become more requent, but these arerare in most settings.

6. Percentage of inpatients with a discharge diagnosis of malaria, and

7. Percentage of inpatient deaths due to malaria

Tese indicators may also be examined to assess trends in malaria. Like test positivity rates, they are less sensitive to changes in reporting rates and health acility use rates. A disadvantage o these indicators is that changes in attendance or conditions other than malaria can aect thepercentage o inpatients with a discharge diagnosis o malaria and deaths due to malaria, e.g.non-malaria inpatient cases and deaths in children < 5 years old could decline i child survivaland immunization activities have rapidly achieved high coverage (e.g. introduction o vaccinationor Haemophilus inuenzae type B, pneumococcus and rotavirus). Furthermore, changes in thepercentages o case and deaths due to malaria (as well as slide positivity rates) will not reect

percentage changes in malaria cases or incidence, as the number o malaria cases is part o thedenominator.

8. Annual blood examination rate

Tis indicator provides inormation on overall diagnostic activity and can be useul in interpretingtrends in malaria cases. While some past guidance suggested that the annual blood examinationrate should be in the region o 10% in order to provide reliable trends, the empirical evidence orsuch a target is not strong. In high-transmission settings, the rate is likely to greatly exceed 10%;2 e.g. in the Solomon Islands, the rate is 60%, just as a result o passive case detection.

9. Percentage of suspected malaria cases receiving a diagnostic test

Programmes should ensure that the percentage o suspected cases receiving a diagnostic test is100% by monitoring the indicator continually, nding out why some health acilities achieve< 100% (e.g. because o RD stock-outs or lack o training) and taking appropriate action.During the period that diagnostic testing is being expanded, it may be dicult to obtain anaccurate picture o trends. It is likely that the number o conrmed cases will increase while thenumber o presumed cases decreases (the number o suspected cases may decrease i guidanceon who should be tested is more restricted). Slide and RD positivity rates might provide some

1 Kahama-Maro J et al. Low quality o routine microscopy or malaria at dierent levels o the health system in Dares Salaam. Malaria Journal , 2011, 10:332.

2 In a population o 10 000, it is expected that about 20%, or 2000, will be under 5 years o age. Each child may have

our episodes o ever per year, o which 40%, or 3200, can be expected to be seen at a public health acility. I allreceive a diagnostic test, the annual blood examination rate will be 32% or child evers alone.

2




inormation on disease trends but could be inuenced by a change in the composition o thepopulation being tested as testing is made more widely available. (It will not be possible to use thepercentage o all attendances due to malaria as an indicator, as the indicator is also aected by the change in diagnostic practice.) Alternatively, it may be possible to explore trends in inpatientcases and deaths or anaemia and blood transusions in children < 5 years, as these are less likely

to be aected by a change in the rate o testing.

1

In such situations, it may be preerable to ocusthe analysis, at least temporari ly, on health acilities in which diagnostic practice has not changedmuch (e.g. health acilities that have always used microscopy), until reliable time series are builtup elsewhere.

10. Completeness of reporting

Programmes should ensure that all health acilities report in a timely manner (usually within2 weeks o a month’s end). Completeness o reporting is usually assessed by the number o monthly reports received rom health acilities in relation to the number expected (the number o healthacilities multiplied by the number o months considered). Tis indicator gives equal weight to allhealth acilities and so may not reect the completeness o case reporting; missing reports rom

district hospitals are likely to represent a larger number o missing cases than missing reports romremote rural health acilities. In either case, the action required is the same: to ollow up missingreports and make arrangements that will acilitate uture reporting. Additionally, completenesso reporting does not take into account the extent to which patients attending health acilities ailto be registered (e.g. because o a large patient load). Special studies are needed to assess the extento this type o problem.

As well as these 10 indicators, it can be useul to examine trends in total outpatient attendanceand total numbers o inpatients and deaths to obtain inormation on overall health acility use,which may be inuenced by changes in user ees, opening o roads or political instability; thesecan strongly aect observed trends in malaria.

2.5 Limitations of surveillance data

Cases reported to malaria surveillance systems represent an incomplete sample o all patientswith ever or malaria. In most malaria-endemic countries, less than hal o all ever cases attendpublic health acilities (rom which the majority o malaria surveillance reports are derived). Tedata reported may also be incomplete because not all ebrile patients receive a diagnostic test orbecause health acilities do not register all patients or submit monthly reports. Tese actors canmake comparison o incidence rates among areas dicult (see Box 2.3).

Not only do health acility surveillance reports represent only a raction o all ever and malariacases occurring in the community, but they may also represent a biased sample, in that healthacility attendees may live closer to the acilities and have better access to medicines and a rangeo government services and economic opportunities. Tereore, it is possible that trends observedin health acilities are not representative o broader trends in the community. Te extent to whichbias occurs is inuenced by the overall accessibility o services: bias is more likely to occur whenonly a small percentage o patients with ever or malaria are treated in the health system. Bias isalso inuenced by the extent to which interventions that aect malaria incidence are correlatedwith the presence o health acilities. As implementation o some interventions, such as wide-scaledistribution o insecticide-treated nets (INs) or indoor residual spraying, is not linked to healthacilities, reductions in the number o cases are just as likely to occur in populations that aredistant rom health acilities as those who live close by.

1 In low-transmission areas, the probability that malaria is the underlying cause o inpatient anaemia in children may be low and hence may not be relevant in monitoring malaria.




Box 2.3.

Influence of health facility attendance, diagnostic testing and reporting rates onreported malaria incidence rates

While incidence rates derived rom surveillance o malaria cases take into account the size

o the population, they may not always reect the true incidence o malaria in the populationbecause:

• Te proportion o suspected cases that attend public health acilities (rom which mostdata are derived) may dier by area and over time.

• Te proportion o people attending public health acilities who receive a diagnostic testmay dier by area and over time.

• Health acility reporting rates may dier by area.

Te example below shows two districts, one urban and one rural, with dierent rates o malaria in the community. Te incidence in the urban district is hal that in the rural district,

but a larger proportion o patients seek care in public health acilities, a larger proportionreceive a diagnostic test and a larger proportion o health acilities submit monthly reports.As a consequence o all these actors, the reported incidence o malaria is higher in theurban district (14 per 1000) than in the rural one (12 per 1000).

Urban district Rural district

a t b c 1,000 50 100

B % c d bc c 60% 40%

C dcd 1,000 (a * B) 30 40

C % d c dc 60% 50%

C dcd 1,000 (a * B * C) 18 20

D % c 80% 60%C dcd 1,000 (a * B * C * D) 14 12

pc c dcd 29% 12%

Tus, it is sometimes observed that areas with better access to and better health acilitieshave a higher incidence o malaria than areas with limited access. It is thereore useul tolook at other indicators (overall health acility use rates, percentage o cases receiving adiagnostic test, completeness o health acility reporting) to help interpret data. It may alsobe useul to examine other indicators, such as diagnostic test positivity rates.

I the acility use rates and reporting rates are known, incidence rates derived rommalaria cases seen in health acilities can be adjusted or these actors, to provide a more

representative estimate o incidence (see World malaria report 2008. Geneva, World HealthOrganization, 2008, Annex 1).

Te likelihood o bias can be explored by examining the results o nationally representativehousehold surveys, such as demographic and health surveys, ‘multiple indicator cluster surveys’or malaria indicator surveys. Tese can indicate the extent to which people with ever use publichealth acilities, private providers or stay at home. Household surveys rom around the worldsuggest that about 40% o people with ever typically seek treatment in public health acilitiesalthough the proportion varies between countries and within countries. Not all cases o everrecorded in household surveys will be malaria, however, and it is possible that people with

non-malarious ever are less likely to seek treatment in health acilities. Household surveyscan also provide inormation on the extent to which interventions such as IN distribution areimplemented equitably (by wealth quintile or urban or rural area).

2




Malaria programmes should also estimate the proportion o all deaths that occur in healthacilities, by comparing the number o deaths recorded in health acilities with the total numbero deaths expected to occur in a country or area (see Box 2.4). Tis will give some indication o the completeness o death reporting. Even so, the proportionate breakdown o deaths by causeobserved in health acilities may not be representative o all deaths that occur in the community.

Box 2.4.

Calculating the percentage of deaths that occur in health facilities

Te number o deaths occurring in a country is calculated by multiplying the total populationby the rate at which people die (the crude death rate). Similarly, to calculate the number o deaths occurring in any age group, the population size o each age group is multiplied by the rate at which people in that age group die (age-specic death rate). Population sizes by age group are usually available rom statistics bureaux in countries. Age-specic death ratesmay be available rom the same source or can be derived rom lie tables produced by WHO.Sample calculations or Zambia in 2010 are shown below. Tese indicate that approximately 20% o deaths occurred in health acilities overall, with slightly higher rates or people over

5 years o age.

Populationa Age specficdeath ratesb

Epectednumber of

deaths

No. of deathsin healthfacilitiesc

% of deathsoccurring in

facilities

A B A*B C C/(A*B)

0-4 2,412,000 0.03086 74,434 14,370 19%

5+ 8,972,000 0.00876 78,595 18,990 24%

Total 11,384,000 0.01344 153,029 33,360 22%

a From United Nations population prospects 2010. http://esa.un.org/unpd/wpp/unpp/panel_indicators.htm .b WHO lie table. http://www.who.int/healthino/statistics/mortality_lie_tables/en/index.html .c

From Ministry o Health, Zambia.

When possible, the results rom surveillance systems should be compared with other sources o data to veriy interpretations or explore why dierences occur. It is o particular interest to examinegeographical variation or trends over time in parasite prevalence and the prevalence o anaemiain children under 5. Tese indicators are measured in an increasing number o household surveysin high- and moderate-transmission settings (see Box 2.5). Nationally representative parasiteprevalence surveys are generally less useul when nationwide parasitaemia has declined to < 5%,as measurement errors, seasonality and geographical heterogeneity can make the results dicultto interpret.1 In these settings, the requirement or large sample sizes and the high associated

costs ofen preclude implementation o such surveys.

1 Hay SI, Smith DL, Snow RW. Measuring malaria endemicity rom intense to interrupted transmission. Lancet Inectious Diseases, 2008, 8:369–378.




Box 2.5.

Comparing surveillance data with household surveys

Zambia reported downward trends in malaria inpatient cases among children under 5 yearso age between 2004 and 2008 but in 2009 there were small increases in 5 o 9 provinces

and large increases in Eastern and Luapula provinces. Household surveys undertaken in2006, 2008 and 2010 showed similar increases in parasite prevalence in Eastern and Luapulaprovinces. An increase in parasite prevalence was also noted in Northern Province. Apartrom dierent measures being used, dierences may also be due to the latest parasiteprevalence survey being conducted in 2010 whereas the latest year or which surveillancedata were available when the analysis was conducted was 2009.

M a l a r i a i n p a t i e n t s p e r 1 , 0

0 0

Malaria inpatients <5 years per 1,000

2006 2008 2009

C e n t r a l

C o p p

e r b e

l t

E a s t e r n

L u a p

u l a

L u s a k a

N o r t h

W e s t e r n

N o r t h

e r n

S o u t h e

r n

W e s t e r n

100

120

40

60

80

0

20

% c

h i l d r e n i n

f e c t e d

Parasite prevalence <5 years

C e n t r a l

C o p p

e r b e

l t

E a s t e

r n

L u a p

u l a

L u s a k a

N o r t h

W e s t e r n

N o r t h

e r n

S o u t h e

r n

W e s t e

r n

50

60

20

30

40

0

10

2006 2008 2010

While surveillance data are subject to incomplete reporting and bias, this source o inormation

has the advantage o continuous collection rom every district in a country; or most districts,such data are the only readily available source o inormation on malaria that can be used by programme managers. When surveillance systems are working well, they show consistent seasonal

variation in the numbers o cases, coinciding with the pattern o malaria transmission. Tey alsoshow reductions in morbidity and mortality in response to interventions and can alert managersto unexpected increases. Tus, although results should be interpreted with care, they are a criticalsource o inormation or programme management and should not be ignored. Continued eortsto improve reporting systems and use o these data will help to improve the quality o malariasurveillance and the operation o malaria control programmes.

2.6 Using surveillance dataObjectives of surveillance systems

Malaria surveillance systems have two principle objectives, to provide programme managers withinormation to identiy geographical locations and population groups in which the incidence o malaria cases and deaths are greatest and to track changes in the incidence o malaria cases anddeaths over time.

Identiy geographical locations and population groups in which the incidence o malaria cases and deaths are greatest. Inormation on where malaria incidence is greatest helps programmes todirect resources to populations in greatest need, e.g. to make sure that sucient INs are availableto populations living in high-transmission areas or that laboratory acilities are upgraded in

areas with the most suspected cases. Both the absolute numbers o malaria cases and deaths and

2




the rate per 1000 population are relevant in deciding where resources should be preerentially allocated. On the one hand, it is important to tackle the problem in populations where the risk isgreatest; interventions can help to reduce morbidity and mortality in these populations markedly.On the other hand, it is important to ensure that programmes conront the bulk o the problem,as judged by absolute numbers. In high- and moderate-transmission settings, there is generally

less geographical heterogeneity than in low-transmission settings; consequently, there may bemuch less scope to target resources dierentially by area, and malaria control interventionsmay be implemented evenly across a district. As control measures are introduced, however, any heterogeneity in incidence may become more pronounced, reecting variation in implementingcontrol measures (e.g. gaps in distribution o INs, lack o IN use, insecticide resistance) asopposed to environmental actors.

Follow changes in the incidence o malaria cases and deaths over time. Inormation on changes inmalaria incidence over time can help programme managers to assess whether their interventionshave been successul in reducing cases and deaths and can help to detect outbreaks that may require special responses. Programme managers in high-transmission areas will be particularly interested to observe whether the numbers o cases and deaths are being reduced or whether

problems are being experienced in some locations and, as a consequence, the programme shouldbe modied. Areas with high malaria transmission are generally not prone to epidemics, althoughthere may still be marked seasonality in the occurrence o malaria and changes rom year to yearin the intensity o transmission owing to climatic actors. Managers should be aware o and beprepared or such uctuations.

Change in emphasis as malaria programmes progress

In the initial phase o malaria control, attention is usually ocused on strengthening surveillancesystems, in particular ensuring improvement in two indicators, namely: the percentage o suspected cases that receive a diagnostic test and completeness o reporting. It may not bepossible to undertake all the analyses on malaria morbidity and mortality desired; e.g. it may be

dicult to examine trends or an entire district, and attention might have to be conned to healthacilities that report consistently, until reliable data are obtained rom all acilities. Nevertheless,it is important to try to improve data quality and to review data whenever possible in order toidentiy problems in implementation and, when possible, accelerate progress in malaria control.

As malaria control programmes are scaled up and transmission declines, the epidemiology o malaria is likely to change in the ollowing ways:

•Te numbers o severe and hospitalized cases and deaths decrease markedly.

•Te number o uncomplicated conrmed cases decreases.

•Malaria transmission becomes more ocal.

•Te age distribution o cases, severe cases and deaths shifs to older children and adults.

•Populations become less immune, the risk or epidemics increases, and these may be responsibleor signicant atalities in such populations.

• Imported cases may represent an important raction o the overall incidence.

Improved health inrastructure and changing epidemiology demand a change in the approach tosurveillance. As malaria becomes more ocal and concentrated in particular population groups,more attention and analysis o indicators by health acility or population groups is needed totarget resources more precisely. Malaria may be concentrated in marginalized populations, suchas those living in remote border areas, migrant workers and tribal populations, and programmesshould nd innovative ways to reach these groups.




In low-transmission settings, data must also be reviewed more requently at health acility levelin order to detect outbreaks as soon as possible. Epidemics may be more likely in areas in whichmalaria has been controlled successully but in which ecient vectors remain than in areas thathave had low levels o transmission due to environmental actors or inecient vectors. Managersshould be alert to malaria outbreaks and be ready to intensiy control measures in some locations

in order to prevent or contain outbreaks.Case investigation and reporting of individual cases

In the initial phase o control, it is recommended that each severe malaria case and death beinvestigated at health acility level, with the support o district sta, to identiy and addressprogramme weaknesses (such as poor coverage with INs, delays in seeking treatment andstock-outs o antimalarial medicines). As transmission is reduced and the number o severe casesdecreases, the opportunities or intensiying investigation o severe cases and deaths increase.It becomes possible to establish a district-wide register o all severe cases, with the aim o investigating and eliminating uture cases and addressing programme weaknesses.

As transmission decreases urther, malaria programmes at the district level can begin to establish

registers o all conrmed malaria cases reported in the district. Te registers can containinormation on the background characteristics o each case (e.g. location, age, sex, occupationalgroup). Analysis o such registers can help to identiy which population groups are most aected,in order better to target interventions and urther accelerate malaria control.

Heterogeneity in programme implementation

Malaria control may progress more rapidly in some parts o a country than in others, and thestrategy or surveillance may thereore vary; e.g. some districts may rely exclusively on reportingaggregate cases, while others may supplement this with reporting the details o individual cases.Indeed, some parts o a country might be pursuing elimination and thereore have to identiy the origin o each case in order to intensiy control measures in specic localities to ensure that

transmission is halted at the earliest possibly opportunity. Countries with a preponderance o low-transmission areas may wish to examine the manual on Disease surveillance or malariaelimination to see i some strategies can be adapted or use in the control phase.

2




3. D cd, , d

3.1 Recording

Communities and health posts

A register should be kept by community health workers and health posts that records, or eachattendance, the date o attendance, patient’s name, village o residence, sex, age, whether it is anew attendance or repeat visit or the same episode o illness and the malaria test result, diagnosisand treatment given (Annex 4). Such inormation will allow community health workers or sta

at health posts to identiy the epidemiological characteristics o malaria in their area (such asthe age and sex breakdown o cases and the locations in which most cases originate). In low-transmission settings, travel history and work location may help identiy sources o inection. Teregister should indicate any cases that are subsequently reerred.

Health centres and hospitals

Outpatients. A register should be kept at health acility level or each outpatient attendance,which records the date o attendance, patient’s name, residence, sex, age, whether it is a newattendance or repeat visit or the same episode o illness, initial diagnosis, type o malaria test,test result, nal diagnosis and treatment given (Annex 4). Tis inormation will enable sta atthe health acility to identiy the epidemiological characteristics o malaria in their area (such

as the age and sex breakdown o cases and the locations in which most cases originate). In low-transmission settings, travel history and work location may help identiy sources o inection.Te register should include cases that are subsequently admitted; attempts should also be madeto include inpatients who bypass the outpatient department, so that a complete record is kept o all cases attending the health acility. As outpatient registers are used or all outpatients, and not

just people with malaria, the existing registers may have to be modied to allow or collectiono this inormation, by the addition o columns or changing column headings. Additionalinormation, beyond that routinely collected in outpatient registers, will be needed or malariacase investigation and reporting o individual cases to district level (see Box 3.2).

Practices or registration o cases vary widely by country. In settings where malaria cases

comprise a high proportion o all outpatients, a separate malaria register will not be kept. Ideally,the diagnosis recorded in the outpatient register would be the nal diagnosis afer administrationo a parasitological test (microscopy or RD); however, in some large health acilities, theoutpatient register may list the health condition initially suspected by a health worker in anoutpatient department (suspected malaria), while malaria tests are undertaken in a laboratory and the results are recorded in a separate laboratory register (conrmed malaria). Te treatmentgiven is sometimes recorded in a hospital pharmacy register (or not at all i patients have topurchase drugs rom pharmacies outside a hospital). In such cases, a monthly report may includeaggregate numbers or the initial diagnosis (suspected malaria) as well as test results (numbertested, number o conrmed cases) and malaria treatment given. While individual patient resultsare not tracked, there should be correspondence between the number o suspected cases, the



21Data reCorDing, reporting, analysis anD use

number o conrmed cases and the number o treatments given. In low-transmission settings,a dedicated malaria notication register may be used to record individual case details. Malariacase investigators should collect pages rom the register weekly, a rst duplicate orm being orelectronic data entry, a second duplicate or case investigation and the original staying in thehealth acility.

ally sheets can also be kept in health acilities to help in calculating the relevant statistics, duringthe day, at the end o the day or, in less busy clinics, at the end o the week (Annex 5). Te tally sheets should be consistent with the requirements o reporting to district level. Daily totals o malaria cases should be recorded in a book, keeping track o daily attendances or all major causes(the categories being be the same as those on the monthly reporting orm used in a country). Atthe end o the week, the daily totals should be tallied to provide a weekly total, and at the endo the month the weekly totals should be tallied to provide monthly totals (Annex 6). Such tally sheets and daily records should be kept or all conditions reported monthly, not just malaria.

Inpatients. Practices or registration o inpatients vary by country. Health acilities may maintainadmission, ward and discharge registers. Te most important one or malaria surveillancesystems is the discharge register, which contains the nal diagnosis.1 Discharge registers shouldcontain the date o admission, the patient’s name, residence, age, sex, diagnosis, length o stay andreason or leaving (discharged, died, transerred, absconded) (Annex 7). Tis inormation shouldbe abstracted rom the patient’s le by appropriately trained sta. I a separate discharge registeris not compiled, sta may abstract this inormation on malaria discharges specically. In smallerhealth acilities, abstraction may be done by the ocer in charge, while in larger acilities thismay be undertaken by a medical records clerk, consulting with attending physicians i necessary.In health acilities that undertake disease coding, malaria diagnoses should ollow the practicesoutlined in the relevant version o the International Classication o Diseases and identiy thespecies o parasite (see Box 3.1). Note that discharge registers are used or all inpatients and not

just those with malaria; additional inormation will be needed or malaria case investigation andreporting o individual cases to district level.

1 In disease outbreaks, admission registers may be used to provide more timely inormation.

3




Box 3.1.

Malaria classification in the International Classification of Diseases (ICD), revisions9 and 10

ICD10 codes

B50 Plasmodium alciparum malaria

B51 Plasmodium vivax malaria

B52 Plasmodium malariae malaria

B53 Other specied malaria

B54 Unspecied malaria

ICD9 codes

084 Malaria

084.0 Falciparum malaria [malignant tertian]

084.1 Vivax malaria [benign tertian]

084.2 Quartan malaria

084.3 Ovale malaria

084.4 Other malaria

084.5 Mixed malaria

084.6 Malaria, unspecied

084.7 Induced malaria

084.8 Blackwater ever084.9 Other pernicious complications o malaria

District level

eams at district or intermediate level should maintain the ollowing inormation:

•monthly reports o the number o suspected malaria cases, number o malaria tests perormedand number o conrmed cases, with total outpatient attendances, inpatients and deaths. Inlow-transmission settings, weekly reports may be kept.

•

a register o malaria programme health structures and sta (health acilities by type andpersonnel by specialty: nurse assistants, nurses, clinicians, laboratory technicians, parasitologists,epidemiologists and their proessional qualications). Te register should be updated annually and include a list o all health acilities and laboratories undertaking malaria testing.

•maps showing the distribution o conrmed cases, inpatients and deaths by health acility catchment area, village or administrative boundary, to be updated annually; and

•all reports and analyses produced by district sta during the past 5 years and submitted tohigher levels. Feedback and other inormation rom higher levels should be collated.

National level

Databases held at national level should have the same structure as those at district level. Whenpossible, inormation should be recorded by health acility (rather than aggregated by district)to allow a greater range o analyses, although this may be challenging in countries with largenumbers o health acilities.




3.2 Reporting

Reports from health posts and community health workers to health facilities

I a health acility supervises health posts or community-based providers, the ollowing datashould be reported monthly to the supervising health acility (Annex 8):

•number o suspected malaria cases seen,

•number o suspected malaria cases tested with an RD,

•number o conrmed malaria cases < 5 years o age (positive RDs),

•number o conrmed malaria cases aged > 5 years (positive RDs),

•number o conrmed cases treated with antimalarial medicine,

•number o presumed malaria cases (not tested) treated with antimalarial medicine, and

•number o cases reerred to a higher-level acility.

Te summed values or these indicators, the number o community health workers expected

to report and the actual number reporting can be written on the health acility reporting orm.Te data should be kept separate and not added to health centre attendances in order to avoidaecting trends over time by uctuations in reporting rom lower-level acilities (e.g. a suddenoutbreak o cases may be assumed i several late reports are received rom health posts).

Reports from health facilities to the district level

High-transmission settings: Each month, health acilities in high-transmission settings shouldreport to the district level the numbers o:

• suspected malaria cases,

•cases tested by microscopy,

•cases tested by RD,

•cases conrmed by microscopy (< 5 and > 5 years o age),

•cases conrmed by RD (< 5 and > 5 years o age),

•conrmed cases treated with antimalarial medicine,

•presumed malaria cases (not tested) treated with antimalarial medicine,

• inpatient cases o malaria (< 5 and > 5 years o age), and

•deaths rom malaria (< 5 and > 5 years o age).

Annex 9 gives examples o data elements to be included on report orms. In areas where testingo all suspected cases is standard practice, it might be assumed that it is unnecessary to reportsuspected cases, as all suspected cases will be tested. esting might, however, not always be possible,because laboratory sta are o duty or there is a stock-out o RDs; hence, it recommended tomaintain the capacity to report total suspected cases. Te numbers o suspected and tested casesshould include patients admitted or malaria (i.e. the numbers o discharges, deaths, transersand absconders). Patients tested with both an RD and microscopy should be counted only once,either or RD or microscopy. In areas with a mix o species, a species breakdown should bereported or conrmed cases (number o P. alciparum, P. vivax, P. malariae, P. ovale, P. knowlesi,and mixed inections).

3




It is useul to obtain inormation on the total numbers o new outpatient attendances (< 5 and> 5 years o age), inpatients (< 5 and > 5 years o age) and deaths (< 5 and > 5 years o age). Tisinormation should be readily available, as in high transmission areas it is expected that malaria

variables would be contained in an integrated health inormation system.

Te number o reports submitted by health acilities to districts and other levels should be kept to

a minimum, although there may some advantage in sending duplicate orms to both district andcentral level simultaneously to speed up the ow o inormation.

Low-transmission settings: Te same three outpatient data elements (suspected, tested, conrmed)and the same two inpatient data elements (cases, deaths) or at least two age groups (< 5 and> 5 years o age) should be reported, with total numbers o outpatients, inpatients and deaths (< 5and > 5 years o age) (see Annex 9).

As the requency o severe cases and deaths decreases, line lists o inpatient malaria cases andinpatient deaths can be reported to district level (Annex 10), with the intention o intensiyinganalysis and response; these represent severe outcomes that should not occur and should be a highpriority or elimination.

Health acilities can move to monthly reporting to the district o line lists o all conrmed malariacases (Annex 11), when the average number o conrmed malaria cases at health acilities is 5–10per month or ewer (in districts with 20 health acilities, data entry o 100–200 records per monthwould be required; in a country with 1000 health acilities reporting 5–10 cases per month, thenumber o cases would be 5000–10 000 per month or 60 000–120 000 per year). Health acilitiesand districts would continue to report aggregate data on the numbers o suspected, tested andconrmed cases in order to calculate the ull range o indicators outlined above.

Core variables or line listing o conrmed cases should include both demographic and clinicalactors (date o attendance, patient’s name, village or suburb o residence, sex, age, type o test, testresult) and intervention actors (IN ownership and use in past 14 days, house sprayed, treatment

received) (see Annex 11). Optional variables could include: date o onset o symptoms, date o rst contact with the health system, presence o gametocytes on microscopy slides (P. alciparum only), occupation, other intervention actors, travel outside the country, travel outside the districtto endemic areas within the country and other data that are locally or nationally important.Even basic inormation on individual cases, such as village o residence, is, however, helpul;programmes should avoid being overly ambitious in collecting additional case inormation so asto ensure complete, accurate reporting.

3.3 Data analysis

Health facility level

Data should be reviewed at least monthly in order to answer the questions below. In hospitalsin which sta have little involvement in preventive programmes, sta rom the district malariaprogramme team may assist in this analayis.

How can severe case and deaths be prevented? In order to undertake initiatives that will have thegreatest impact on malaria morbidity and mortality, health acility sta should ocus on malariainpatients and malaria-related deaths and investigate, at the individual level, why they acquiredmalaria and how it developed into severe disease (see Box 3.2). Investigation o cases shortly afer admission to wards may enable health acility sta to recognize gaps in the provision o preventive measures (INs or indoor residual spraying), the use o preventive measures, delays inseeking treatment or poor access to diagnostic testing and treatment. Te gaps identied can be

addressed by making better use o health acility resources or liaising with people at community level or other government services.




Box 3.2.

Investigation of inpatient cases and deaths

Data on each inpatient case o conrmed malaria (a proxy or severe malaria) or malariadeath should be used to investigate possible programme weaknesses in the prevention or

treatment o malaria.

1. Examine the age, pregnancy status, village and month o illness to characterize theepidemiology o the admitted cases and deaths. It is important to assess whether casesarise in particular villages, to plan a more ocused response, or are scattered throughoutthe health acility catchment area, suggesting that a more generalized response isnecessary.

2. Examine the preventive measures used:

•Did the household o the patient own a long-lasting insecticidal net (LLIN)? Yes, no

•Are more than two people using each LLIN? Yes, no

•Were LLINs hanging during the 2 weeks beore hospitalization? Never, some nights,all nights

•Did the patient sleep under an LLIN in the 2 weeks beore symptoms o malariadeveloped? Never, some nights, all nights

• I he or she did, how old was the LLIN? < 1, 1–3, > 3 years

• I he or she did, what was the condition o the net? Holes < 100 cm² (good), holes> 100 cm² (too torn) or holes that can be repaired (not too torn, serviceable).

I ew patients report owning an LLIN (e.g. < 25%), LLIN ownership is a major problem in thehealth acility catchment area. I a high proportion o patients report owning an LLIN but many

report not using them, LLIN usage is major problem in the health acility catchment area. I ahigh proportion o patients used an LLIN (e.g. > 50%1), the ecacy o the LLIN may be in doubt(e.g. they may have holes, the insecticide may have deteriorated or mosquitoes may be developingresistance to the insecticide). Further entomological investigation may be required.

3. Examine the treatment received.

•Was treatment received? Yes, no.

•Was treatment received promptly afer onset o symptoms? < 24 h, 1–3 days, > 3 days.

• I treatment was received, were the medicine and dosage appropriate? Yes, no.

•What were the reasons or delay in receiving treatment? Not aware o treatment,

distance to acility, stock-outs, cost o medicine.

In order to answer these questions, it is necessary to record: the date o onset o symptoms; dateo rst contact with the health system; diagnostic test administered, i any; test result; type o antimalarial medicine received; date o rst dose o antimalarial medicine; and date o admission.

In most situations, treatment with artemisinin-based combination therapy is highly eective and should prevent uncomplicated malarial inection rom proceeding to severemalaria. I a high proportion o severe malaria cases have not received this therapy, it may be necessary to determine ways o improving access to and compliance with treatment. I,however, a signicant proportion o cases (e.g. > 25%) report taking a ull 3-day course o artemisinin-based combination therapy, the possibility o low ecacy o the medicines (e.g.

ake or substandard medicines or drug resistance) should be considered.1 As the ecacy o INs is not 100%, isolated cases o malaria may be expected in users o LLINs.

3




Are there unusual changes in the numbers o cases? Te number o conrmed malaria cases shouldbe plotted monthly to identiy any variation over time. Tis assessment can be made more easily i the previous 2 or 3 years o data are plotted on the same chart. Te latest month’s value orconrmed cases can be compared with an ‘85th percentile’ threshold value to determine whetherit is unusually low or high (see Box 3.3). Te 85th percentile threshold is derived by examining the

previous 3 years o monthly case numbers and calculating the fh highest number o cases thatoccurred in the past 3 years.1 I the latest month’s value exceeds the 85th percentile, district levelsta should be notied, and an investigation should be undertaken to determine whether urtheraction is necessary. I the number o cases remains stable despite an increase in malaria controlinterventions, such as recent distribution o INs, an investigation may also be merited to assesswhether the INs have reached the target populations, are being used and are eective against thelocal anopheline vector population.

Box 3.3.

Calculating an 85th percentile threshold to assess whether the numbers of casesare higher than usual

Te monthly number o conrmed malaria cases in the current year should be plotted on agraph with the previous 2 or 3 years o data to determine whether or not there are variationsin the numbers o cases. Te latest month’s value or conrmed cases can be compared withan 85th percentile threshold value to determine whether it is unusually low or high. Te laest2 months’ values in the example exceed the 85th percentile and thereore look unusually high.

J F MA M J J A S OND J F MAM J J A S O ND J F MAM J J A S OND J F MAM J

Confirmed malaria cases

85th percentile

N u m b e r o f c o n f i r m e d

m a l a r i a c

a s e s

0

50

100

150

200

250

300

350

400

450

2008 2009 2010 2011

1 In 3 years there are 36 months. Te 85th percenti le is the fh highest value in the 36 months, as 15% o the monthly values wi ll be greater than or equal to this (36 * 15% = 5).




Box 3.3. ContinueD

Calculating an 85th percentile threshold to assess whether the numbers of casesare higher than usual

Te 85th percentile threshold is derived by examining the previous 3 years o monthly case

numbers and calculating the fh largest number o cases that occurred in the past 3 years.In the example below, the previous 36 monthly values were:

Number of confirmed cases

2008 2009 2010 2011

Jan 195 217 263

Feb 193 156 244

Mar 231 162 224

Apr 205 196 289

May 242 237 310

Jun 372 301 299 411

Jul 310 221 249

Aug 251 216 210

Sep 145 151 152

Oct 199 188 76

Nov 184 175 103

Dec 188 158 118

Current month Four highest values in previous 3 years Fifh highest value in previous 3 years (85th percentile)

Te 85th percentile is the fh highest value within the 36 months, as 15% o the monthly values will be greater than or equal to this (36 * 15% = 5).

Graphs o numbers o inpatients and deaths may also reveal trends, i there are sucient numbers.Graphs can be particularly useul i diagnostic testing practices have changed over time. Inthese settings, trends in the total number o conrmed cases should be assessed along with thenumber o patients receiving a diagnostic test; it is dicult to assess trends in the total numberso conrmed cases. An additional graph by month showing the test positivity rates at the healthacility and observed by community agents should be prepared monthly i the data are available.

Do some areas have more malaria than others? Annually (or more requently), it is useul to

determine which vil lages are reporting the most cases; even in high-transmission settings, malariacan be concentrated in particular areas in which the environmental conditions are particularly suitable or transmission.1 Health acility records o malaria cases are inuenced by the distancepatients have to travel to the health acility, so that more cases are generally rom within 5 kmo a acility than rom urther away. It is thereore useul to review data rom health posts andcommunity agents, i they are available. I an unusually large number o cases arise in one villagein relation to the population served, it might be useul to review the availability o preventivemeasures and access to treatment and undertake measures to improve them i necessary.

1 In Kenya, 20% o houses with parasites in the dry season yielded 65% o the cases in the ollowing wet season.

Bejon P et al. Stable and unstable malaria hotspots in longitudinal cohort studies in Kenya. PLoS Medicine, 2010,7:e1000304. doi:10.1371/journal.pmed.

3




Low-transmission settings: In low-transmission settings, where there is a greater risk or epidemicsor greater annual variation in the number o cases, the numbers o suspected, tested and conrmedmalaria cases should be plotted weekly in order to detect unusual variations at an early stage. Forweekly plots, the 85th percentile is obtained by examining the previous 3 years o weekly casenumbers and calculating the 23rd highest number o cases that occurred in the past 3 years.1 I

the latest week’s value exceeds the 85th percentile, the district level sta should be notied, and aninvestigation should be undertaken to determine whether urther action is necessary (Box 3.2).

District level

Data should be reviewed at least monthly in two ways: or the district as a whole and or individualhealth acilities or geographical areas:

1. Trends in malaria should be examined throughout the district . Tis will enable managers toanswer the ollowing questions:

•Are testing and reporting targets being met; e.g. what percentage o suspected cases are tested,and what percentage o health acilities have submitted reports in a timely ashion?

•Are there trends in malarial disease that are o concern; e.g. a rise in the number o cases oran unusually large number o deaths in 1 month that suggest that control activities should beintensied.

•Are there unusual dierences between indicators; e.g. does the number o deaths remainconstant despite the number o cases alling?

In order to perorm these analyses, districts should update ve surveillance graphs every monthto monitor trends in malaria cases and deaths (see Box 3.4). Data should be presented or thecurrent year and the previous 3 years:

1 In 3 years there are 156 weeks. Te 85th percentile is the 23rd highest value within the 156 weeks, as 15% o monthly values wi ll be greater than or equal to this (156 * 15% = 23).




Box 3.4.

Eamining trends in a district

Confrimed cases per 1000

Inpatients per 10 000

Deaths per 100 000

Annual blood examination rate

Slide positivitiy rate

% Inpatients due to malaria

% Deaths due to malaria

% Health facilities reporting

% Suspected cases tested

% Cases due to P. falciparum

Outpatients per 1000

Inpatients per 10 000

Deaths per 100 000

0%

20%

40%

60%

80%

100%

0%

10%

20%

30%

40%

50%

0%

1%

2%

3%

4%

-

50

100

150

200

250

-

0%

20%

40%

60%

80%

-

5

10

15

20

2009 2010 2011 2012

2009 2010 2011 2012 2009 2010 2011 2012

2009 2010 2011 2012

2009 2010 2011 20122009 2010 2011 2012

1. Malaria incidence rates

2. Proportional malaria incidence

3. General patient attendance

4. Diagnostic effort

5. Quality of diagnosis and reporting

6. % Cases due to P. falciparum

Each month, districts should update ve control charts showing trends in malaria incidenceand other indicators (or six i more than one species o malaria is present). In the examplepresented, there has been a reduction in the numbers o malaria cases, inpatients and deathsin recent months, particularly at the beginning o 2012 (chart 1). Tis decrease appears to berelated to lower outpatient attendance and admissions overall (chart 3) and a reduced rateo diagnostic testing (chart 4), which could also be due to lower reporting rates in recentmonths (chart 5). Te slide positivity rate and percentages o inpatients and deaths due tomalaria do not appear to ollow the downward trend in the number o malaria cases (apartrom a dip in early 2012) (chart 2). Similarly, there has been no marked changed in thepercentage o cases due to P. alciparum (chart 6).

Hence, there appears to be no real decrease in the number o malaria cases; the apparentdecrease is due to less reporting in recent months. Such a pattern, in which data areincomplete or the most recent months, is common in many reporting systems and suggeststhat eort is needed to improve the timeliness o reporting (otherwise, the reporting rates arereasonably good at > 90%). Tere is also scope to increase the percentage o suspected casesreceiving a diagnostic test, and it might be necessary to determine why this has decreased inthe most recent months (perhaps due to the selection o health acilities reporting).

3




(i) Malaria incidence rates. Tis chart shows trends in the numbers o conrmed malaria casesper 1000 population, malaria inpatients per 10 000 population and deaths per 100 000 population.It may be necessary to experiment with the size o the reerence population (1000, 10 000 or100 000) so that the lines t conveniently on the same graph (or they may be plotted on a secondaxis or separate graphs). Tese indicators are basic measures o morbidity and mortality that

reect the success o preventive programmes and indicate demand or treatment in the publicsector.

(ii) Proportional malaria incidence. Tis chart shows the positivity rate with slides and/or RDs,the percentage o inpatients with a discharge diagnosis o malaria and the percentage o inpatientdeaths due to malaria. Tese measures are aected less by uctuations in reporting rate andtotal patient attendance and sometimes more reliably indicate the direction o change in malariamorbidity and mortality. Changes in test positivity rates and proportionate malaria cases ordeaths (proportion o all-cause cases or deaths) do not, however, reect the percentage change inmalaria cases or incidence, as the number o malaria cases is part o the denominator. In areaswhere the percentage o deaths due to malaria is small and uctuates erratically, it may not beuseul to plot the percentage o deaths due to malaria.

(iii) General patient attendance. Tis chart shows the total number o outpatients per 1000population, the total number o inpatients per 10 000 population and the number o deaths per100 000 population. It may be necessary to experiment with the size o the reerence population(1000, 10 000 or 100 000) so that the lines t conveniently on the same graph (or they may beplotted on a second axis or separate graphs). Tese indicators provide inormation on overall useo health acilities, which can inuence observed trends in malaria. Tis chart will also be useulor interpreting other disease-specic data that are collected and analysed regularly.

(iv) Diagnostic activity. Tis chart shows the annual blood examination rate, which reects totaldiagnostic activity in a population. Tis can inuence observed trends in malaria and is thereoreimportant contextual inormation.

(v) Quality o diagnosis and reporting. Tis chart shows the percentage o suspected malaria casesreceiving a diagnostic test and the percentage o health acilities submitting reports each month.Te target or both indicators should be 100%. Tis can provide inormation on the extent o parasitological diagnosis and the completeness o reporting, both o which can inuence observedtrends in malaria.

(vi) Percentage o cases due to P. alciparum. In countries in which P. vivax and P. alciparum are present, an additional chart showing the percentage o cases due to P. alciparum should bedrawn.

2. Indicators or diferent health acilities or geographical areas should be compared. Tis willenable managers to answer the ollowing questions:

•Which health acilities are testing and reporting adequately and which are experiencingproblems? For example, are some health acilities unable to increase the percentage o suspectedcases tested?

•Are there unusual dierences between health acilities or some indicators? For example, is thenumber o cases not decreasing despite an increase in IN coverage?

Such comparisons can be made by three methods: by examining control charts or each healthacility, by constructing surveillance tables or each indicator or by constructing a summary tableo surveillance indicators by health acility with trend statistics.




Examining control charts or each health acility: Tis might take a prohibitively long time ordistricts with many health acilities; charts or health acilities that require additional attentionshould be examined selectively.

Constructing surveillance tables or each indicator. Tese tables are similar to the control chartsdescribed above:

•able 1: Counts o total conrmed cases, inpatients and deaths by health acility and month.Tese tables can be maintained as wall charts and updated each month (see Box 3.5). Forconstruction o the table and because dierent proportions o cases o ever attend each healthacility, it is easier to record absolute counts o cases rather than incidence rates.

Box 3.5.

Eamining trends in health facilities: counts of total numbers of confirmed cases,inpatients and deaths by health facility and month

Wall charts should be kept at the district oce showing monthly numbers o conrmedcases, inpatients and deaths by health acility in order to track unexpected changes and

identiy health acilities experiencing particular problems. Te examples show seasonality in the incidence o malaria with possibily higher than normal number o cases in healthcentre M in October.

Confirmed cases Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

h a 730 613 389 341 419 398 380 294 628 557

h B 277 275 152 173 176 122 88 66 159 302

h c a 2 0 0 18 0 15 2 2 0 0

h c B 11 0 0 0 0 0 0 0 5 18

h c C 97 0 67 0 0 0 102 95 137 92

h c D 24 18 8 20 14 6 9 11 19 25

h c e 13 15 5 3 11 6 2 0 0 0h c f 0 53 23 0 78 10 12 11 23 20

h c g 1 0 0 0 0 0 0 0 0 0

h c h 1 0 0 4 3 0 0 0 0 0

h c i 0 3 5 3 1 4 7 10 0 0

h c J 41 43 17 12 28 24 29 4 15 40

h c k 2 1 6 4 4 0 1 2 3 5

h c l 10 1 10 79 111 8 7 0 0 0

h c m 7 0 0 0 0 0 5 4 24 21

h c o 12 12 13 9 4 4 2 4 6 13

Total 1228 1034 695 666 849 597 646 503 1019 1093

3




Box 3.5. ContinueD

Eamining trends in health facilities: counts of total numbers of confirmed cases,inpatients and deaths by health facility and month

Inpatients Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

h a 12 15 9 4 12 4 7 3 31 6

h B 33 13 3 2 2 2 39 4 2 28

h c a 1

h c B

h c C 1 1 2 3 11 26

h c D 3 1 1 1

h c e 1

h c f 13 10 5 2

h c g

h c h 1

h c i

h c J 1 1

h c k

h c l 1 2

h c m 1 1

h c o

Total 49 42 24 7 22 8 48 9 45 61

Deaths Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

h a 1 1 1 3

h B 1 1 2

h c a

h c B

h c C

h c D

h c e 1

h c f

h c g

h c h

h c i

h c J

h c k 1

h c lh c m

h c o

Total 2 1 2 2 2 3

•able 2: Microscopic slide and/or RD positivity rate and percentage o inpatients with adischarge diagnosis o malaria. Given the low requency o malaria deaths, it is generally notuseul to show the percentage o deaths due to malaria by health acility.

•able 3: Annual blood examination rate, percentage o suspected cases tested, health acility report submitted or not (or date submitted or number o days delay in submission).

•able 4: otal numbers o outpatients, inpatients and deaths.




Such charts allow inspection o the results o all health acilities and identication o unusualtrends. When examining the results, it might be helpul to calculate a summary trend statisticor each health acility to determine whether the latest month’s result is unusual. Tis is relatively straightorward to do i control tables are maintained in an electronic spreadsheet but may becumbersome i wall charts are maintained by hand.

Four possible trend statistics, in increasing order o complexity, are:Latest month versus a reerence value: Te latest month’s results can be compared with a baseline

value, such as the same month 5 years previously. Tis procedure can be useul in assessingprogress towards targets in a national health plan, e.g. to reduce cases by hal between 2010 and2014 or since commencement o control activities (see Box 3.6 or application to district data).

Box 3.6.

Eamining trends in health districts: summary table of surveillance indicators byhealth facility with trend statistics

A table o surveillance indicators by district, with trend statistics, can used to examine

trends in several districts simultaneously. Over short periods with relatively small changesin population size, counts o the number o cases are likely to be adequate or assessment.Over longer periods (> 5 years), when population growth has a greater eect on populationsize, incidence rates are preerable. rends in completeness o reporting, percentage o suspected cases tested and annual blood examination rates can be used to assess whetherchanges in reporting or diagnostic practice are responsible or any changes in incidence or i the trend in case incidence is likely to be real. In the example below, there are large decreasesin the number o malaria cases, case incidence and test positivity rate. Te annual bloodexamination rate and the percentage o suspected cases have decreased since the baselineperiod, but these cannot explain the ull change in incidence or in slide positivity rate. Teinvestigation should begin by determining the reasons or the decreases in the percentage o

suspected cases tested and the annual blood examination rate.

3




Box 3.6. ContinueD

Eamining trends in health districts: summary table of surveillance indicators bydistrict with trend statistics

Current indicator values

Change from 5 years

previously1 (%)

R e g i o n / D i s t r i c t

P o p u l a t i o n ( 1 0 0 0 s )

N o . o f h e a l t h f a c i l i t i e s

C o n f i r m e d c a s e s

I n c i d e n c e r a t e / 1 0 0 0

p o p u l a t i o n

T e s t p o s i t i v i t y r a t e ( % )

A B E R

T e s t e d / s u s p e c t e d ( % )

R e p o r t i n g c o m p l e t e n e s s

( % )

C o n f i r m e d c a s e s

I n c i d e n c e r a t e / 1 0 0 0

p o p u l a t i o n

T e s t p o s i t i v i t y r a t e ( % )

A B E R 2

T e s t e d / s u s p e c t e d ( % )

R e p o r t i n g c o m p l e t e n e s s

( % )

Province ADc 1 287 12 22,747 79 28 28 76 97 -19 -23 -44 -37 -8 -3

Dc 2 342 15 28,199 83 24 34 82 100 18 12 -54 -142 -1 17

Dc 3 249 12 20,681 83 24 34 83 100 -7 -12 -56 -98 -28 4Dc 4 272 13 5,788 21 16 13 64 99 -48 -51 -76 -107 24 10

Dc 5 270 13 6,126 23 9 27 71 100 -73 -74 -86 -84 -19 18

Dc 6 318 17 49,358 155 34 46 72 100 125 114 -49 -320 14 15

Dc 7 250 11 21,194 85 26 33 70 100 -30 -34 -62 -72 -30 0

sub total 1,986 93 154,093 78 25 31 75 100 -4 -9 -57 -114 -10 9

Province BDc 1 388 10 9,647 25 15 17 80 100 -61 -70 -66 13 -35 0

Dc 2 240 6 8,595 36 13 28 87 100 -49 -34 -59 -60 -58 9

Dc 3 241 6 4,230 18 6 31 75 100 -79 -71 -82 -58 -55 0

sub total 869 22 22,472 26 11 24 80 100 -64 -61 -69 -27 -49 3

Province CDc 1 312 11 2,938 9 8 12 94 91 -81 -82 -82 -4 -16 -9

Dc 2 326 20 2,707 8 4 23 96 100 -91 -93 -92 18 -22 18

Dc 3 351 21 5,582 16 18 9 89 74 -42 -39 -51 -25 41 -8

Dc 4 323 12 759 2 3 9 75 94 -95 -96 -94 39 -44 2

Dc 5 258 16 5,959 23 10 22 88 100 -72 -75 -79 -21 -19 1

sub total 1,570 80 17,945 11 8 15 90 93 -80 -83 -82 2 -17 2

Province DDc 1 271 13 38,816 143 34 42 80 100 -25 -28 -41 -21 -22 5

Dc 2 281 13 22,567 80 20 41 85 100 -48 -51 -60 -23 -18 0

Dc 3 281 11 8,184 29 16 18 94 100 -72 -73 -74 -2 33 1

Dc 4 292 14 1,972 7 5 13 80 98 -94 -94 -89 46 -9 -2

Dc 5 304 14 1,503 5 6 9 69 100 -76 -77 -81 -21 -36 0

Dc 6 256 13 18,491 72 25 29 80 92 -10 -15 -46 -59 -40 1

Dc 7 250 13 1,367 5 6 9 76 85 -89 -90 -88 14 -3 -14

Dc 8 274 14 7,976 29 13 22 76 99 -74 -76 -76 0 -21 12

sub total 2,209 105 100,876 46 20 23 81 100 -55 -58 -61 -8 -15 4

Province E

Dc 1 286 18 3,633 13 7 19 80 100 -74 -75 -82 -37 -25 0Dc 2 304 9 3,613 12 16 7 61 100 0 -5 -65 -169 -6 0

Dc 3 291 12 769 3 5 5 79 100 -60 -62 -85 -154 -32 12

Dc 4 335 17 1,946 6 4 14 72 89 -89 -90 -87 16 -37 -10

Dc 5 340 9 402 1 2 8 79 100 -85 -86 -94 -127 -49 24

Dc 6 348 12 9,611 28 15 19 77 100 -71 -72 -67 17 -5 0

Dc 7 289 16 875 3 4 8 79 100 -81 -82 -90 -75 -4 29

sub total 2,193 93 20,849 10 8 11 75 100 -73 -74 -78 -19 -20 8

Total 8,828 393 316,235 36 18 20 79 100 -49 -51 -63 -32 -21 7

1 Negative numbers indicate decrease2 ABER: Annual blood examination rate




Latest 3 months versus previous 6 months: A weighted total o the values or the latest 3 months iscalculated in relation to that or the previous 6 months.

(Xt-2 + 2* Xt-1 + 3*Xt )

(Xt-5 + Xt-4 + Xt-3 + Xt-2 + Xt-1 + Xt )

where Xt = latest month’s value, Xt–1 = previous month’s value, Xt–2 = value 2 months earlier, etc.I this ratio is larger than 1.1, it can indicate that the results or the past 3 months are unusually high in comparison with a recent reerence period. A ratio lower than 0.9 can indicate that theresults or the past 3 months are unusually low (dierent thresholds can be explored, e.g. 1.2 or0.8, depending on the sensitivity required). Te statistic is most useul in situations in whichshort-term changes are considered worthy o attention; it is less useul or indicators that areexpected to uctuate seasonally.

Latest month versus 85th percentiles or previous 3 years. 85th percentile thresholds are derived by examining the values o an indicator or the previous 3 years and calculating the fh highest valuethat occurred in the past 3 years (Box 3.3).1 Te value or the latest month might be considered

unusually high i it is higher than the 85th percentile.Latest month versus similar months in previous years. Te latest month’s value is compared withan average o values or the same month in the previous 5 years, the previous months in theprevious 5 years and the ollowing month in the previous 5 years. Tus, i the latest month isApril, an average o results or the months March to May in the previous 5 years is calculated,with the standard deviation. I the latest month’s value is greater than or lower than one standarddeviation rom the mean, the value may be unusually high or low in relation to the previous5 years. Te statistic may be useul or detecting outbreaks. 2

Constructing a summary table o surveillance indicators by health acility with trend statistics. Atable showing values or indicators in the latest month, with a trend statistic to indicate changesthat have occurred, can be constructed, as shown in Box 3.6.

Programme managers may not be able to explore all the options or data presentation. At the very least, the control charts or the district as a whole should be updated each month, together withthe control tables showing the numbers o malaria cases, inpatients and deaths by health acility,completeness o reporting and the percentage o cases receiving a diagnostic test.

Low-transmission areas

Districts in which severe cases and deaths are increasingly rare can intensiy their analysis andresponse to inpatient malaria cases and deaths. Every inpatient malaria case and all malariadeaths should be investigated by location, age, season, parasite species and intervention actors(possession and use o LLINs, indoor residual spraying, type and timing o treatment) to identiy

weaknesses in the public health system and to understand why the severe cases were not preventedor adequately treated (see Annex 10 or the type o data to be reviewed). District managers shouldalso consider whether some villages are not covered by the public health system and may still beexperiencing severe cases and deaths (by examining the extent to which all vi llages are representedin outpatient registers).

As the incidence o cases decreases urther and districts can build up a register o all cases, it willbe possible to examine the risk actors associated with malaria inection (e.g. age, sex, location,occupation) in order that programmes can be more closely targeted to the populations mostaected (see Annex 11 or the type o data to be reviewed).

1 In 3 years there are 36 months. Te 85th percentile is the fh highest value within the 36 months, as 15% o monthly values will be greater than or equal to this (36 * 15% = 5).

2 eutsch SM, Churchill RE. Principles and practice o public health surveillance. Oxord, Oxord University Press,2000.

3




National level

National-level programmes should conduct analyses similar to those in districts: (i) analysis o overall trends nationally on the ve control charts, (ii) comparison o districts rom the controltables and (iii) comparison o districts rom summary tables o surveillance indicators. Inaddition, districts can be compared using maps (see Box 3.7).

Box 3.7.

Eamining trends in health districts: using maps

Mapping o indicators allows programme managers to assess whether programmeperormance or malaria trends vary by geographical area and to determine whether malariaprevention, testing or treatment activities should be ocused in particular geographicalareas. Regional dierences in the numbers o cases and deaths due to malaria might reectthe underlying epidemiology, the extent o malaria interventions or diagnostic and casereporting practices. In the example below, higher case incidence rates are observed ineastern parts o the country, with higher annual blood examination rates and percentages o cases tested. Nonetheless, a higher incidence rate is suggested by higher test positivity ratesin the same areas. Variation in the completeness o reporting may be due to communicationdelays or resource gaps in particular regions.

Incidence of conrmed malaria cases Malaria test positivity rate

Annual blood examination rate

Rate of completeness of reporting by health facilities

Percentage of suspected cases tested

<10

10 - 40

>40

Conmred MalariaIncidence Rate per 1000 pop

<10

10 - 20

>20

ABER

<10

10 - 20

>20

Test Positivity rate

<50

50 - 80

>80

% Suspected cases tested

<50

50-80>80

Reporting rate




Analysis is most likely to be done quarterly and annually, as reports are submitted by districts atdierent times. Te aims are to ascertain whether targets are being met, whether there are trendsin disease that are o concern, which districts are reducing the number o cases and which areexperiencing problems and any unusual dierences between indicators (e.g. ailure o the numbero cases to decrease despite an increase in IN coverage). When necessary, individual district

control charts can be inspected to explore issues in more detail, and data can be retrieved romindividual health acilities.

3.4 Using data for making decisions

Formal meetings. I the data generated by a surveillance system are to be used to improve theoperation o malaria control programmes, programme managers must make sure there areregular opportunities or review. A schedule o meetings should be established to review malariatrends, such as the ollowing:

•community with health acility sta: monthly or quarterly

•health acility sta with district malaria control programme sta: monthly

•district sta with national malaria control programme sta: quarterly. Meetings might have toheld less requently or held regionally in order to create opportunities or national sta to meetwith all district sta during a year.

Supervision. Supervision rom national and district level is needed to support building o theinormation system, ensure the completeness o reporting, ensure analysis and discussion o dataand ollow-up o recommended actions. During visits to health acilities and district team oces,supervisors should check that registers are kept up to date, with all elds completed, that data onreport orms correspond to the inormation in registers and tally sheets, that core analysis graphsand tables are up to date and that discussions are held about interpretation o the trends andpotential action (see Annex 12 or an example o a malaria surveillance supervisory checklist).Health acility sta should be encouraged to investigate all inpatient malaria cases and deaths.

Feedback. District managers should prepare eedback or health acilities monthly or quarterly,including private health acilities that provide data. Tis should not simply reect the datasubmitted by the health acility but should include comparisons with other health acilities in thedistrict and summary statistics or the district as a whole. A regular bulletin can be produced ina standard ormat to present district results (based on control charts) and comparisons o healthacilities.

A national eedback bulletin should be produced each quarter, showing indicators by district(Annex 13). As transmission is reduced, mapping could be extended to subdistricts in orderto present more detailed epidemiological inormation on remaining locations and population

groups aected. Te bulletin should be widely circulated, not only as eedback to districts, butalso as inormation or other government departments, institutions and implementing partners.Elected leaders should also be presented with the bulletin on malaria, possibly showing themalaria situation according to political boundaries, to instil understanding and support ormalaria control at the highest level o leadership.

3




4. eb c c

Surveillance systems consist o the tools, procedures, people and structures required to generateinormation or planning, monitoring and evaluating malaria programmes.

•Te tools include: report orms, tally sheets, registers, patient cards, computer hardware andsofware, documentation and training materials.

•Procedures include: case denitions, reporting requency, pathways o inormation ow, dataquality checks, incentive schemes, data analysis, mechanisms or review o perormance,

methods or disseminating results, using data or making decisions, supervision and planning.•Te people include: decision-makers both inside and outside the health service who use data

rom surveillance systems, the health sta who gather or use the data and the community whose details are registered.

•Te structures include the ways sta are organized to manage, develop and use the system.

Deciencies in any o these components can limit the capacity o a malaria control programmeto undertake disease surveillance eectively. Tereore, the building o a unctioning, sustainablesurveillance systems must usually address each o these areas.

4.1 ToolsSuggestions or registers and orms are included in this manual, which should be adapted or useby countries. Registers should include space to record essential data elements, such as test results,and the inclusion o additional variables should be avoided. Te more variables used in registersand orms, the less likely the orms will be completed accurately, i at all. When possible, ormsshould reect existing guidance, such as that provided in standard treatment manuals.

In the control phase, malaria surveillance systems are ofen part o communicable diseasesurveillance or health inormation system. Health inormation systems should be adapted toinclude the basic data elements suggested in this manual (Annex 9). It is important to involveall stakeholders during discussions about revising systems, especially those involved in data

collection at health care level, who may provide valuable input into the constraints aced by healthworkers and practical suggestions or moving orward. An inclusive process also creates greaterownership and encourages the adoption and use o orms later. In low-transmission settings,especially where malaria is relatively rare and conned to particular locations, there may be aseparate malaria reporting system (allowing a timely response to individual cases o malaria),which should be adapted according to the recommendations in this manual.

I new orms are designed or existing orms are revised, they should be tested on a small scale (e.g.one district or 6 months) beore widespread implementation. Afer nal adjustments have beenmade, documentation on use o the orms should be updated and data collectors trained in theiruse. As training will be required in each district, changes cannot be implemented at one time butmust be staggered over the course o a year. It is usually advantageous to link training in malaria

surveillance systems with other training activities in order to save costs and to avoid occupying toomuch o health workers’ time. When possible, training in malaria surveillance should accompany training on malaria case management, particularly the use o diagnostic testing.



39estaBlishing surveillanCe systems in the Control phase

As new orms are introduced, old ones should be removed rom health acilities or destroyed,so that health workers do not use previous systems when there are disruptions to supplies o stationary or because they are not amiliar with the new orms. In order to maintain consistentsupplies o stationary, the cost o orms might be borne centrally rather than by individualdistricts. It might also be useul to liaise with local printing companies to ensure that they are

amiliar with the latest orm designs and do not print old orms when a district submits an order.Pre-service curricula o medical and nursing schools should also be updated to reect the latestdata recording and reporting practices.

4.2 Procedures

Procedures or recording, collating, storing, securing, cleaning, querying and analysing datashould be documented. In addition, procedures or reporting and eeding back inormationshould be documented, so that all sta using the system are clear about their responsibilities.Particular attention should be paid to ensuring that all health acilities report accurately andin a timely ashion. Programmes should have an up-to-date inventory o all health acilities

that are expected to report and should ollow up delays. District sta should undertake regularsupervision o health acilities to ensure that surveillance activities are being conducted properly and that control activities are driven by analysis o surveillance data.

4.3 People

Adequate human resource capacity is required to run and maintain malaria surveillance systems.Particular attention should be given to the time and training required. District malaria ocalpoints need sucient time to conduct analyses and ollow up 10–50 health acilities each month.In addition, the move to case-based analysis requires that more person-time will be spent on

village-level analyses and analyses by age, sex, occupation and other risk actors. At national

level, sucient person-time is needed or data acquisition rom health inormation departments;importing, merging, cleaning and analysing data; mapping; and producing a national eedback bulletin. I sucient resources are available, dedicated personnel or malaria monitoringand evaluation should be placed at all levels (district, regional and national) to coordinate allmonitoring and evaluation and ensure that the tools, procedures and competent, adequately trained personnel are in place. Disease surveillance requires epidemiological, statistical andcomputer skills; district sta need specic training in data analysis by health acility and analysiso case-based data. raining may be institutionalized in an annual workshop or district malariaprogramme managers or surveillance ocers, with the aim o conducting training, reviewingprogress in implementing surveillance systems, reviewing the latest data and preparing work plans or the coming year.

4.4 Structures

Structures or implementing surveillance systems vary by country. In some countries, dataunctions are undertaken by a single health inormation unit rather than being split amongseparate programmes. Tis arrangement can ensure greater coordination in system design andreduce duplication in requests or data. Malaria programme managers must liaise closely withhealth inormation sta to ensure prompt access to relevant data. In other countries, much datamanagement is undertaken by programme sta. In these cases, there should be coordinationwith inormation units to ensure use o common data sets, such as population projections, healthacility lists and coding systems. Opportunities should be created to undertake consolidated

analysis o inormation with other programmes, so that progress in malaria control can be put

4




into perspective. Eorts should be made to coordinate system developments across programmes,perhaps by establishing a ‘health inormation system development committee’ with representationrom a variety o heath programmes and senior management. Such a committee could ensurethat a ministry o health prepares a coherent strategy or developing inormation systems and nota system that is incompatible, unnecessary or unsustainable.



a






43annex 1. types of malaria DiagnostiC test

ANNEx 1.

t dc

Parasite-based diagnostic testing for malaria

Tis section presents a brie summary o inormation on malaria diagnostic testing. More detailedinormation can be ound in Universal access to malaria diagnostic testing: an operational manual 1 and the other documents cited below.

For the management o symptomatic cases, RD and microscopy are usually sensitive enough atany level o malaria transmission. In some settings asymptomatic P. alciparum inections may

persist at very low parasite densities, below the threshold o detection by microscopy or RDs,and can be detected only by molecular methods such as PCR.

Microscopy

Light microscopy has been the standard or malaria diagnosis or many decades and is stillthe primary method o malaria diagnosis in health clinics and hospitals throughout the world.It is the only widespread method o dierentiating between all major Plasmodium species,P. alciparum, P. vivax, P. malariae and P. ovale, as well as or detecting gametocytes o P. alciparum and mixed inections. Microscopy can provide parasite counts (i.e. estimates o parasite density in peripheral blood) and can thereore be used to monitor response to treatment. Microscopy requires unctioning equipment, regular provision o laboratory supplies, well-trained laboratory

technicians at all levels, regular supervision and a unctional quality management system. For thisreason, it is generally more widely available in countries with more resources and more robusthealth systems. In areas where microscopy is not available, especially in high-burden countries,RDs are becoming increasingly available as the standard or malaria testing in outpatientsettings.

Te sensitivity and specicity o light microscopy are directly related to the time availableto read a blood lm, the quality o the stained lm and the competence o the microscopist.Good microscopists in health acilities can detect 100–200 parasites per microlitre, and expertmicroscopists can detect 50 parasites per microlitre. In most endemic areas, nearly all clinicalillness truly due to malaria is thought to correspond to > 100 parasites per microlitre; thereore, agood microscopist should detect parasites in nearly all true clinical malaria cases.2 Occasionally,

clinical cases may occur at lower parasite densities, particularly very early in the course o aninection.

WHO has suggested competence levels or microscopists, with those at the expert level expected toachieve 90% detection, 90% correct species identication and a high level o quantication (to bewithin 25% o the true value 90% o the time), and acceptable levels o clinical competence belowthis value.3 In low-transmission settings, high specicity is vital but is hard to maintain. Field

1 Universal access to malaria diagnostic testing: an operational manual . Geneva, World Health Organization, 2011.http://whqlibdoc.who.int/publications/2011/9789241502092_eng.pd .

2 Parasitological conrmation o malaria diagnosis: report o a WHO technical consultation. Geneva, World Health

Organization, 2010.3 Malaria microscopy quality assurance manual . Manila, WHO Regional Oce or the Western Pacic, 2009.

ANNEx

1




microscopy standards are ofen low, and rigorous quality management systems are necessary tomaintain sucient perormance or both malaria case management and surveillance.

Rapid diagnostic tests

Several RDs currently on the market can consistently detect over 95% o parasite inections

at 200 parasites per microlitre, with 95% specicity. Tere is growing experience o nationwidedeployment o RDs. Recently, batch-to-batch quality issues have been raised, emphasizingthe need or post-procurement lot testing at regional centres established by WHO1 and regularassessment o RD perormance against expert microscopy at health acilities.

WHO has published the results o RD perormance evaluations against panels o wild-typeparasites diluted at specic densities and assessed or stability at high temperatures and easeo use.2 Te evaluations show wide variation in the perormance o dierent products, andprocurement should be undertaken in the light o good evidence. An interactive guide designedto help national malaria control programmes select malaria RDs with specic perormancecharacteristics is available.3 raining, supervision and perormance evaluations o health workersusing RDs are also a necessary part o RD programmes.

Te three main groups o antigens detected by RDs are:

•histidine-rich protein 2, which is specic to P. alciparum;

•Plasmodium lactate dehydrogenase (pLDH), currently used in products that includeP. Òalciparum-specic (pLDH-P), pan-specic (pLDH-Pan), present in all human malariaspecies, and P. vivax-specic pLDH (pLDH-Pv), and non-alciparum specic (pLDH-vom)antibodies; and

•aldolase, which is pan-specic.

Dierent products on the market have dierent combinations o antibodies that can detect theabove antigens. RDs that detect both alciparum-specic and non- alciparum (or pan-specic)

target antigens are commonly called ‘combination’ or ‘combo’ tests. Te commonest ormatso RD products are a plastic cassette and dipsticks; cassettes tend to be simpler to use thandipsticks and have been deployed on a wider scale.

PCR tests

New methods or routine PCR-based surveillance o malaria inections are being used orresearch and eld studies, which are more sensitive than light microscopy or RDs in detectingsubmicroscopic inections, especially with rare species (P. malariae, P. ovale and P. knowlesi),mixed inections and low-density inections. In Cambodia, or example, in a national survey in 2007 in which the populations o 76 villages were screened, 13 more villages with malariacases were identied with PCR than with microscopy.4 During screening and treatment in Pailin,

Cambodia, in 2008–2009, use o PCR with eedback and treatment o positive cases made itpossible to treat 86 asymptomatic carriers (P. vivax in most cases) among the 928 people screened,instead o six when only RDs were used (S. Hoyer, personal communication).

1 Malaria Rapid Diagnostic est Evaluation Programme, World Health Organization. http://www.wpro.who.int/

sites/rdt/who_rdt_evaluation/lot_testing.htm .2 Results o WHO product testing o malaria RDTs: round 3 (2010–2011). Geneva, World Health Organization, 2011.

http://www.who.int/tdr/publications/documents/rdt3.pd .3 Foundation or Innovative New Diagnostics. Malaria RDT product testing: interactive guide. http://www.

nddiagnostics.org/programs/malaria/nd_activities/product_testing/malaria-rdt-product-testing/4 Parasitological conrmation o malaria diagnosis. Geneva, World Health Organization, 2009.



45annex 1. types of malaria DiagnostiC test

Te relation between the incidence o symptomatic malaria and the prevalence o symptomaticinections in a population (called the ‘reservoir’) is not ully understood. It depends partly on theprevalence o low-density inections: the lower the overall parasite prevalence in a population,the more additional inections will be ound by PCR than by microscopy.1 It also depends on thespeed at which malaria transmission decreases: when the decrease in transmission is more rapid

than loss o immunity in a population, the reservoir o asymptomatic carriers can be signicant,and mass screening is potentially appropriate. For example, in Cambodia, microscopy suggesteda 3% prevalence, whereas PCR resulted in a prevalence o 7%. When transmission has decreasedover many years, however, most people with parasitaemia are symptomatic because they haveno immunity, and the reservoir is minimal; e.g. a prevalence o 0% by microscopy, 0.5% by PCRor P. alciparum and 1.5% by PCR or P. vivax in the Brazilian mountains outside Amazonia;2 a prevalence o 0% by PCR in two districts in Sri Lanka.3 In this situation, mass screening willprobably not be cost-eective.

Te potential programme value o detecting low-density inections that are microscopy-negativebut PCR-positive is unclear.

Quality management systems for microscopy and RDT Quality assurance o both microscopy and RDs is an essential component o programmesto strengthen and expand parasitological conrmation o malaria diagnosis. Data on cases o malaria conrmed by either microscopy or RD in countries without unctioning laboratory quality assurance systems may not be reliable. Variations over time in the implementation o laboratory quality management systems may also inuence trends in the numbers o conrmedmalaria cases (or malaria test positivity rates), and, with coverage o laboratory services, must betaken into account in data interpretation. Further guidance on quality management systems isavailable in the reerences below.4,5

1 Okell LC et a l. Submicroscopic inection in Plasmodium alciparum-endemic populations: a systematic review andmeta-analysis. Journal o Inectious Diseases , 2009, 200:1509–1517.

2 Cerruti C Jr. et al. Epidemiologic aspects o the malaria transmission cycle in an area o very low incidence in Brazil. Malaria Journal , 2007, 6:33.

3 Rajakaruna RS et al. Pre-elimination stage o malaria in Sri Lanka: assessing the level o hidden parasites in thepopulation. Malaria Journal , 2010, 9:25.

4 Malaria microscopy quality assurance manual . Version 2. Geneva, World Health Organization, 2009.

5 Universal access to malaria diagnostic testing: an operational manual . Geneva, World Health Organization, 2011.http://whqlibdoc.who.int/publications/2011/9789241502092_eng.pd .

ANNEx

1




ANNEx 2.

D 1

In a patient with P. alciparum, asexual parasitaemia and no obvious cause o symptoms, thepresence o one or more o the ollowing clinical or laboratory eatures classies the patient assuering rom severe malaria:1

Clinical features

• impaired consciousness or unrousable coma;

•prostration, i.e. generalized weakness so that the patient is unable to walk or sit up withoutassistance;

• ailure to eed;

•multiple convulsions: more than two episodes in 24 h;

•deep breathing, respiratory distress (acidotic breathing);

•circulatory collapse or shock, systolic blood pressure < 70 mm Hg in adults and < 50 mm Hgin children;

•clinical jaundice plus evidence o other vital organ dysunction;

•haemoglobinuria;

•abnormal spontaneous bleeding; or

•pulmonary oedema (radiological)

Laboratory findings

•hypoglycaemia (blood glucose < 2.2 mmol/l or < 40 mg/dl);

•metabolic acidosis (plasma bicarbonate < 15 mmol/l);

• severe normocytic anaemia (Hb < 5 g/dl, packed cell volume < 15%);

•haemoglobinuria;

•

hyperparasitaemia (> 2% or 100 000/μl in low-intensity transmission areas or > 5% or 250 000/μlin areas o high, stable malaria transmission intensity);

•hyperlactataemia (lactate > 5 mmol/l); or

• renal impairment (serum creatinine > 256 μmol/l).

1 Guidelines or the treatment o malaria, 2nd Ed. Geneva, World Health Organization, 2010:35.

NNEx

2



47annex 3. Core surveillanCe inDiCators for malaria Control

ANNEx 3.

C c dc c

1. Confirmed malaria cases (number and rate per month or per year)

Formula 1000 * nb cd c

p

Numerator t b cd cd b cc rDt. t bd cd b d c. t b d cd c dcd (d c b c ) c ( c); d bd c dcd d c.rd , -bd (cc rDt), c cc , d b cdd (cd)

c .Denominator t b cc. a

c b dd c c cqd, cd c d. p z d b djd ccd jc c , b, ud n p D jc.

i c d d c c c.

i ccd , , z d bddd b 12 b cdc .

Breakdown h d d : a ( , < 5 ), c, c , ( d )

l : s, 5- , dc (, c, c)

e: fc, , c c: d, c (dcd, d, ),dcd

Purpose t d bd d d c .

t dc d -cdc .

Interpretation td c b cd b:

• c : d c c c c c d c cdd .. cd .

• b d (d d, rDt d)

• c c b d

• c c cdc

i d ccd c dc c c cdc c, dc c , bd d b 1000 b d, c b c .

Other t dc d.

ANNEx

3




2. Inpatient malaria cases (number and rate per month or per year)

Formula 10 000 * nb c

p

Numerator t b d dc d. p bcdd b d d b cdd. a c d d -bd (cc d/ rDt) d dc d bd .

m b d cd b d c bd.Denominator t b cc. a

c b dd c c cqd, cd c d. p z d b djd ccd jc c , b, ud n p D jc.

i c d d c c c.


Breakdown a ( , < 5 ), c , ( d )

Purpose t c d. t dc c c , cc ccd d.

Interpretation i c d d d d dc c .

t b c c c d, d b dcd.

t dc - d d- c bd c.

i c, c cd ccd P. falciparum c (ccd d) , c d cc.

td c b cd b:


• c dc cc .. dc dc dc b dd.

• c c b d

• c c cdc .

i d ccd c dc c c cdc

c, dc b d c ,c c c d dc d b 10 000, c b c .




3. Inpatient malaria deaths (number and rate per month or per year)

Formula 100 000 * nb c

p

Numerator C c d c d . a cdd d d d -bd (cc d/ rDt) d d bd .

D d d c bd d b cdd.

Denominator t b cc. a c b dd c c cqd, cd c d. p z d b djd ccd jc c , b, ud n p D jc.

i c d d c c c .


Breakdown a ( , < 5 ), c , ( d )

Purpose t c b d

Interpretation i d d d dc c .

t b d , d d b d c dc, b c d cc d c.

t dc - c d .td c b cd b:


• c dc cc .. dc dc dc b d dd.

• c c b d

• c c b d

i d ccd c dc c c c, dc d b d c , c c c d dc , b 10 000 db d 100 000, c b c .

4. Malaria test positivity rate (RDT and/or blood slide)Formula 1000 * nb cd c

nb c c

Numerator nb c cd cd b cc rDt. t b dcd b d c. t b d cd c dcd (d c b c ) c ( c); d bd c dcd d c. a -bd (cc rDt), c cc , d b cdd cd c.

Denominator t b cd c d

Breakdown t dc (cc rDt), c, c , ( d), dc (, c, c)

Purpose t c d bd d d .p ‘cc’ c d rDt c- bc dd c d.

Interpretation rDt d d c d d d b d . rDt c c d b cd b .

t c c c d b dd dd cc . C d b ,, cc b cd c d.

C d c c c c cdc, b c d.

ANNEx

3




5. Percentage of cases due to P. falciparum

Formula 100 * nb cd P. falciparum c

nb cd c

Numerator nb P. falciparum c cd b cc rDt. t b d cdb d c. md c P. falciparum d b cd P. falciparum.

t b c cd c dcd (d c b c ) c ( c); d bd c dcd d c.

Denominator t b c cd b cc rDt. t b d cdb d c. t b d cd c dcd (d c b c ) c ( c); d bd c dcd d c.

Breakdown t dc (cc rDt), c , ( d ), dc (, c, c)

Purpose t c c d P. falciparum d d d b c

Interpretation C d d b c d c d b djd dd P. vivax c

C d d c, , c cd , c d P. falciparum dc; P. vivax b d

qc c bc c d cdd bc d (z) b c bc q . P. ovale d P. malariae bc q, b .

a b dc P. falciparum b (cc rDt), c d b c d P. falciparum cd b c d d, .. cc d rDt d b d.

6. Percentage of inpatient cases with a discharge diagnosis of malaria

Formula 100 * nb c dc d

t b

Numerator nb d dc d. p bcdd b d d b cdd. a c d d -bd (cc d/ rDt) d dc d bd .

m b d cd b d c bd.

Denominator t b dcd dd ( bcdd b dd b cdd).

t b d cd b d c bd.

Breakdown a ( , < 5 ), c , ( d )

Purpose t c d. p ‘cc ’ c bc dd c d.

Interpretation i c d d d d dc c .

t b c b c d, d c b dcd.

t dc - d d- c bd c.

i c, c cd ccd P. falciparum (ccd

d) d , c d cc.td c b cd b:

• c : d c c d c cdd,.. d d .

• c dc cc .. dc dc dc b dd.

• c dc cd .. dc b d d c.

• c c cdc .

i d ccd c dc c c cdc , dc d b d c , c c d dc d b c d cd, c b c .

C c dc d d c c b c c cdc b c d.




7. Percentage of inpatient deaths due to malaria

Formula 100 * nb d d

t b d

Numerator nb d d. a c d d -bd (cc d/ rDt) d d bd .

t b d d cd b d c

bd.Denominator t b d

t b d d cd b d c bd.

Breakdown a ( , < 5 ), c , ( d )

Purpose t c b d. p ‘cc’ c bc dd c d.

Interpretation i d d d dc c .

t b d , d d b d c dc, b c d cc d c.

t dc - , c d .

td c b cd b:

• c : d c c d c cdd,

.. d d d .• c dc cc .. dc dc dc

b c dd .

• c dc cd .. dc b d d c.

• c c d .

i d ccd c dc c c d, dc b d c , c c d dc d b d d cd, c b c .

C d c c c b c cdc, b c d.

8. Annual blood eamination rate

Formula 100 * nb c c

p

Numerator t b cd c d. t c cd c d c dc.p d b b rDt d cc d b cd c.

Denominator nb cc. a c b dd c c cqd, cd c d.p z d b djd ccd jc c , b, ud n p D jc.

i ccd , , z d b ddd b12 b .

Breakdown t dc (cc rDt), c , ( d ), dc (, c, c)

Purpose t c dc ; d cdc.

Interpretation h bd c c c.

s dc bd d b b 10% d d b d, b c dc c . i - , cd 10% d c dc .

ANNEx

3




9. Percentage of suspected malaria cases that have had a diagnostic test

Formula 100 * nb c c

nb cd c

Numerator t b cd c d. t d cd d b cdc ; dd b c c dc d b cdd. p d bb rDt d cc d b cd c.

Denominator nb cd c d c (.. c dc).

Breakdown t dc (cc rDt), c , ( d )

Purpose who cd cd c d c dc b cc rDt, d . t dc c c b c d b qd.

Interpretation a 100% dc b d cd, c d, c, c- rDt c cc dc.

10. Completeness of health facility reporting

Formula 100 * nb c cd

nb c cd

Numerator nb cd c. i b c cd b c d d .. c d

d c.i c d b , .. bc cd, z c d c d b cd d d d cd.

Denominator nb c cd. g, b c cd d b b cdd.

Breakdown gc , ( d ), (.. , )

Purpose r d - c c c c . ad c dc.

Interpretation t c c d b 100%. v < 100% dcb , cc, b, .

t dc q c d c c c ; dc cc b c c.

Other i d d dc zd b dc ( b

c), dc d c b: b ccd d b c d.

i c c , c b c d b ccd. t c d c dd d : b c cd db d d .



53annex 4. suggesteD register for Community health workers, health posts anD outpatient Departments

ANNEx 4.

sd c , d d c d

No. Date Name

Residence

(village,

neighbour-

hood) Se

Age in

years

Provisional

diagnosis

New

visit?

Malaria

test

result

Final

diagnosis Treatment

1

2

3

45

6

7

8

9

10

11

12

13

14

15

1617

18

19

20

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)

(6) Age in years: Age should be recorded as < 1 or 0 or children < 1 year o age.

(7) Provisional diagnosis: may be amended in column 11 i the result o a malaria diagnostic testresult is negative.

(9) Malaria test result: Te result should be recorded as +ve, –ve, or not done. I more than onespecies is possible, the parasite species (P.., P.v., P.m., P.o.) should be recorded or positive testresults.

(11) Final diagnosis: Will include presumed malaria i no test was perormed.

(12) reatment: Speciy i artemesinin-based combination therapy or other antimalarial treatmentwas given and i patient reerred.

Te number o suspected malaria cases can be derived rom column 7. Te number o conrmedcases can be derived rom column 9. Te number o presumed malaria cases can be derived by subtracting the number o conrmed malaria cases in column 9 rom the number o malariadiagnoses in column 10. Counts should apply only to new visits, which are indicated in column 8;sometimes, columns or repeat visits are added to the right o column 11.

ANNEx

4




ANNEx 5.

s dc c d

Patient attendance Total

scd ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØooo47

ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo

Microscopy

p d bcc

ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØooo ooooo42


P. falciparum ØØØØØ ØØØØØ ØØØØØ Øoooo ooooo ooooo ooooo ooooo ooooo ooooo 16ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo

P. vivax ØØØØØ Øoooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo6


P. malariae ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo

P. ovale ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo

md ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo

p (cd ) <5

ØØØØØ ØØØØØ ØØooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo12


p (cd ) ≥5

ØØØØØ ØØØØØ ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo10


RDT testing

p d rDt

ØØØØo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo4


p (cd ) <5

Øoooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo1


p (cd ) ≥5



Treatment

Cd cc

ØØØØØ ØØØØØ ØØØØØ ØØØØØ ØØØoo ooooo ooooo ooooo ooooo ooooo22


pd c c (presumed cases = cases not tested)



A tally sheet can be used to make counts rom records in registers or to keep a running total o patients in clinics. Each circle can be viewed as a patient’s head, and a circle is crossed when apatient satises particular criteria. Te tally sheet can be used or daily or weekly totals. At theend o the day or week, the crossed circles are added and the totals transerred to a daily or weekly summary book or chart.

Te tally sheet should be locally adapted. For example, i there is no P. vivax or P. ovale, those canbe removed. In settings where a multi-species RD is used, the RD section should be adaptedto report those results.

NNEx

5



55annex 6. Daily anD weekly reCorDs of outpatient attenDanCe at health Centres anD hospitals

ANNEx 6.

D d cd dc c d

Month: April 2012 1 2 3 4 5 6 7 Weeklytotal

8 9

Day: s m t w t f s s m

Patient attendance

scd 8 59 47

Microscopy

p d 56 42

P. falciparum 18 16P. vivax 3 6

P. malariae

P. ovale

md

p (cd ) <5

8 12

p (cd ) ≥5

13 10

RDT testing

p d rDt 8 2 4p (cd ) <5

2 0 1

p (cd ) ≥5

1 0 0

Treatment

Cd cc

2 0 1

pd c c (presumed cases = cases not tested)

1 0 0

otals rom tally sheets can be copied into a daily and weekly summary book, so that there is apermanent record o the daily counts o outpatient attendance. Tese can be used to assess daily or weekly changes in the incidence o disease and to calculate monthly totals, to be transcribedonto a monthly report. Te order o rows and their height should be the same as those o the tally sheets to acilitate transcription.


ANNEx

6




ANNEx 7.

Dc d c d

No. Date Name

Residence(village,

neighbourhood) Se Age YMD Diagnosis

Lengthof stay(days)

Reason for leaving

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1617

18

19

20

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

(7) YMD: units in which age is recorded: days should be used or children < 1 month, months orchildren < 1 year, years or others).

(8) Diagnosis: Should ollow ICD classications as ar as possible; some acilities may add a

column or the ICD code.(10) Reason or leaving: discharged, died, transerred or absconded

Te total number o malaria inpatient cases should be the number discharged plus died, i.e.excluding transerred and absconded, as a nal diagnosis will not have been made.

NNEx

7



57annex 8. reports from health posts anD Community health workers to health faCilities

ANNEx 8.

r d c c

Patient attendance

scd

Testing

p d rDt

Cd <5

Cd ≥5

Treatment

Cd d dc

C d d dc

C d

Te number o variables to be reported each month should be kept to a minimum, to enhancethe completeness and quality o reporting. All health workers should understand the terms used,i.e. ‘conrmed malaria’: suspected malaria cases with a positive test. Notes can be placed at the

bottom o a orm and in standard treatment manuals as a reminder.

ANNEx

8




ANNEx 9.

r c dc

Areas with P. falciparum only

Outpatients

scd

t

Testing

p d b ccCd <5

Cd 5+

p d rDt

Cd <5

Cd 5+

Inpatients

m <5

m 5+

t <5

t 5+

Deaths

m <5

m 5+

t d <5

t d 5+

Treatment

Cd d dc

C d d dc



59annex 9. reports from health faCilities to the DistriCt level

Areas with more than one species of Plasmodium

Outpatients

scd

t

Testing

p ccc d

P. falciparum

P. vivax

P. malariae

P. ovale

md

t cd <5

t cd 5+

p d rDtCd <5

Cd ≥5

Inpatients

m <5

m ≥5

t <5

t ≥5

Deaths

m <5

m ≥5

t d <5

t d ≥5

Treatment

Cd d dc

C d d dc

Te number o variables to be reported each month should be kept to a minimum, to enhancethe completeness and quality o reporting. All health workers should understand the terms used,i.e. ‘conrmed malaria’: suspected malaria cases with a positive test. Notes can be placed at thebottom o a orm and in standard treatment manuals as a reminder.


ANNEx

9




ANNEx 10.

l c dd b d dc -

(7) ype o test: RD, microscopy or none

(8) Species: I only P. alciparum is present, this column is not needed. I more than one speciesmight be involved, the parasite species (P.., P.v., P.m., P.o.) should be recorded or positive testresults.

(9) IN: insecticide-treated net.

(10) IRS: indoor residual spraying.

(15) Medicines used: Specic details to be provided to determine possibility o expired orcountereit medicines.



61annex 10. line lists of malaria Cases anD Deaths among inpatients to Be reporteD at DistriCt level

M a l a r i a p r e v e n t i o n

A n t i m a l a r i a l t r e a t m e n t

R e a s o n f o r

l e a v i n g

( d i s c h a r g e d /

d i e d /

a b s c o n d e d /

t r a n s f e r r e d )

N o .

D a t e

a d m i t t e d

N a m e

R e s i d e n c e

( v i l l a g e ,

n e i g h b o u r -

h o o d )

S e x

A g e

P r e g -

n a n t ?

( Y / N ) ?

T y p e

o f t e s t

( R D T /

m i c r . )

S p e c i e s

I T N

o w n e

d

b y

h o u s e -

h o l d

( Y / N ) ?

I T N u s e d i n 2

w e e k s b e f o r e

a d m i s s i o n

( a l l n i g h t s

s o m e / n o n e ) ?

H o u s e

r e c e i v e d

I R S

( Y / N ) ?

D a t e o f

o n s e t o f

s y m p -

t o m s

D a t e

c o

n t a c t e d

h e a l t h

s y s t e m

R e c e i e v e d

a n t i m a l a r i a l

t r e a t m e n t

( Y / N )

D a t e

s t a r t e d

M e d i c i n e s

u s e d ( A

C T ,

C Q , o t h e r s )

1 2 3 4 5 6 7 8 9 1 0

1 1

1 2

1 3

1 4

1 5

1 6

1 7

1 8

1 9

2 0

( 1 )

( 2 )

( 3 )

( 4 )

( 5 )

( 6 )

( 7 )

( 8 )

( 9 )

( 1 0 )

( 1 1 )

( 1 2 )

( 1 3 )

( 1 4 )

( 1 5 )

( 1 6 )

( 1 7 )

ANNEx

10




ANNEx 11.

l cd c b d dc -

(7) Species: I only P. alciparum is present, this column is not needed. I more than one speciesmight be involved, the parasite species (P.., P.v., P.m., P.o.) should be recorded or positive testresults.

(8) IN, insecticide-treated net;

(10) IRS, indoor residual spraying; (9)ype o test: RD, microscopy or none

(15) AC, artemisinin-based combined therapy; CQ, chloroquine



63annex 11. line lists of all ConfirmeD malaria Cases to Be reporteD at DistriCt level

M a l a r i a p r e v e n t i o n

A n t i m a l a r i a l t r e a t m e n t

N o .

D a t e

a d m i t t e d

N a m e

R e s i d e n c e

( v i l l a g e ,

n e i g h b o u r -

h o o d )

S e x

A g e

T y p e

o f t e s t

( R D T /

m i c r . )

S p

e c i e s

I T N

o w n e d

b y

h o u s e -

h o l d

( Y / N ) ?

I T N u s e d i n 2

w e e k s b e f o r e

a d m i s s i o n

( a l l n i g h t s

s o m e / n o n e ) ?

H o u s e

r e c e i v e d

I R S

( Y / N ) ?

D a t e o f

o n s e t o f

s y m p -

t o m s

D a t e

c o n t a c t e d

h e a l t h

s y s t e m

R e c e i e v e d

a n t i m a l a r i a l

t r e a t m e n t

( Y / N )

D a t e

s t a r t e d

M e d i c i n e s

u s e d ( A C T ,

C Q , o t h e r s )

1 2 3 4 5 6 7 8 9 1 0

1 1

1 2

1 3

1 4

1 5

1 6

1 7

1 8

1 9

2 0

( 1 )

( 2 )

( 3 )

( 4 )

( 5 )

( 6 )

( 7 )

( 8 )

( 9 )

( 1 0 )

( 1 1 )

( 1 2 )

( 1 3 )

( 1 4 )

( 1 5 )

ANNEx

11




ANNEx 12.

s cc c d

During visits to health acilities, supervisors should check that registers are kept up to date, withall elds completed, that data on report orms correspond to inormation in registers and tally sheets, that core analysis graphs and tables are up to date and that discussions are held aboutinterpretation o the trends and potential action. Health acility sta should be encouraged toinvestigate all malaria inpatient cases and deaths. An example o a supervisory checklist orsurveillance or malaria is shown below.

Record keeping Not presentPresent but not up

to datePresent and up to

datePresent, up to date

and no mistakes

o √

Dc √

D dc b √

m dc b √

g cd c √

g b d √

g b cd c √

g √

Reporting None 1 2 3

nb

3 √

Investigations performed in past 3months

Not done Done Done & actiontaken

m d √

m √

m c √

Disease or progamme delivery issues that need attention

l b c lcd

Recommendations

Cc dd.

w lcd c c d llin d d c .

NNEx

12



65annex 13. moDel monthly Bulletin for Countries with high or moDerate transmission

ANNEx 13.

e q b c d

A national eedback bulletin should be produced each quarter, with data by district. Te bulletinshould be widely circulated, not only as eedback to districts but also as inormation or othergovernment departments and institutions. Elected leaders should also be given the bulletinon malaria, possibly showing the malaria situation according to political boundaries, to instilunderstanding and support or malaria control at the highest level o leadership.

Bulletins can draw on the control charts described in Box 4.4 but should be tailored to country circumstances, e.g. programme priorities or availability o data. In addition to surveillance charts,country bulletins should include some measures o intervention coverage. An example o the rstpage o a country bulletin is shown below (other pages show tables o indicators calculated ordistricts). Te ormat allows the sharing o a large amount o inormation in a small space. Itshould be noted that gures with more than three trend lines may be dicult to interpret.

ANNEx

13









